ADS1299 Normal Electrode mode

Vibodha Bandara

Other Parts Discussed in Thread: ADS1299

I build an circuit board with the great help of EEG Front-End Performance Demonstration Kit and the recommendations on ADS1299 Manual .

I can read and write registers correctly to the chip and it works fine with the internally generated test signal as the image bellow.

But when I set a channel to normal electrode mode the problem starts.. Here I set CHANNEL 1 to normal elactrode mode with Gain 1 and others are in test signal mode. But there is an corrupted value in CHANNEL 8 which repeat every time as follows

Then I change the gain value of the CH1 1 to 12 problem remain same and another problem started at gain =12... the output CH 1 becomes at its lowest value (800000h). this happens for Gain = 12 ,24

Then I Change all the Cannels to Normal Electrode mode and Gain = 24.. all the sample values of the channels gets 800000H except CHANNEL 7 . CHANNEL 7 gets the Maximum binary value(7FFFFFH)...

I have Enable the internal reference and Clock..

Please Help me to solve this problem

over 10 years ago

0 Vibodha Bandara over 10 years ago

Intellectual 260 points

please....

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

What are the inputs connected to?

Regards,
Brian

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Hi Brian, Nice to see you again.. :D

The inputs are not connected to any devices.. They are just open...

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

You're going to need to connect them to an actively driven signal source with a common mode voltage within the range of the input PGA. If the inputs are floating then the output will be undetermined.

Regards,
Brian Pisani

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Hi brainy..

So what about EEG Front-End Performance Demonstration Kit .. It does not show this kind of output when I put floating the inputs..
Is it ok if i connect 10mV signal generator to the inputs ?? If yeas I'll attach it and post the output here ...

Another question... The Default value of the CHnSET register is 60H ok?? but here I get it as 61H which is in input short mode... is it ok? or is there any problem ??

Thank you...

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

Floating the inputs is not a deterministic phenomenon so I wouldn't read too much into the differences between you system and the EVM in that regard.

Yes you may connect a signal generator. Just make sure that the generator is biased properly with respect to the device analog supplies. One of the most common mistakes I see is people connecting the signal generator ground to the negative input of the device which is just a floating gate. You must actively bias it.

The datasheet indicates that the POR value of CHnSET is 0x61. Are you seeing otherwise?

Regards,
Brian Pisani

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Hi Brian ,

Actually my plane was to connect the ground of the generator to SRB1 and the signal to on channel by anabling SRB1 bit.. According to you its not correct I guess ?? Please guide me to Bias it correctly.. Is that connecting Generator ground with the ADS1299 ground??
Thanks

Data sheet Page 44 : The default values of the each bit of CHnSET register gives 60H.. yeas page 39 says the reset value as 61H.. I thought "default values"= "reset values"... any way its not a big issue..

Thank you

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

You can connect it to ground, but then you need to provide a DC offset to ensure the signal stays within the range of the PGA. In addition, you'll need a way of providing a signal to the negative input as well, whether it be DC or otherwise. The difficulty is that the signal generator wants to produce a single-ended signal but the converter only converts differentially. It's also possible to tie the negative input and the sig gen ground to the bias output when it is configured as a simple buffer and then provide the signal to the positive input. If you are in bipolar supply mode, you can do this with the board ground.

Regards,
Brian Pisani

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Dear Brian,

Today I connect the device with the signal generator 10 mVpp .. My board is Bipolar.. I connect Negetive_input(CH0) + signal Ground to the BIAS out put.. but the output was not as expected.. Then I add the board ground to BIAS output... Then I got somewhat better graph.. but they are not in good condition.....

CH1 Gain=1

1Hz , 10 mVpp

5 Hz

10 Hz

The internal references is also in 4.5V correctly....

The magnitude of the graphs are also confusing all are in minus values.... ( the vlaues are ADS values without converting to voltages)

Thank you.

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

First, do not connect the bias output to the board ground. They voltages will be close but may not be exactly the same so it will create a short circuit.

If you connect only the negative device input and the signal generator ground to the board ground (no bias), does the output look better?

Regards,
Brian Pisani

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

dear brian..
actually in my board i use SRB1 pin insted of negetive pin....(i cannoth use negetive pin in this board).... so i connected board ground and signal ground to SRB1 but no good thing observed....
thank you

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

Have you set the SRB1 bit in the MISC1 register?

Regards,
Brian Pisani

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Dear Brian,

Today I found a bug in my program which is not set the MISC1 register correctly.. I found this after you asking the register setting.. Thank you very much for it....

.. Now It gives better results for lower frequencies < 5Hz. The waveform is going to deform after 5Hz.. Following results were observed by connecting signal ground to SRB1 pin + Board Ground. Same results observed with using BIAS...

1HZ

5Hz

10 Hz

15 Hz

This is when applying 10 Hz square wave.....

Thank you..

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

Try plotting in Matlab using just points. I believe Matlab may be drawing straight lines in between your samples which is making it look distorted.

Brian

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

hi brian..,
yes that can be true... the how can i check weather it actually distorted or not..
thanks in advance ..

0 Vibodha Bandara over 10 years ago in reply to Brian Pisani

Intellectual 260 points

Hello brian... thank you very much for ur help.. i was able to solve my main problems with your contribution...
Thanking you
Vibodha

0 Brian Pisani over 10 years ago in reply to Vibodha Bandara

TI__Mastermind 38745 points

Hey Vibodha,

I'm glad you've got things working. Let me know if you need anything else.

Regards,
Brian Pisani

Data converters

Data converters forum

ADS1299 Normal Electrode mode