ADS1299 is working but readings are far from correct!

Hey Brian!

My PGA gain is set to 1 so that won't change my code.

I don't know if the channel inputs interfer anything when in input short mode, but anyway... they are grounded.

So, these are the readings I get in all channels (input short):

Some things I noticed:

Channel 8 NEVER has a value of 0.xxxx.

Sometimes other channels get a 4.4xxx value.

Here are register values, just in case.

Btw, arduino won't display MSBs of a word if they are 0.

Hey Brian, sorry for taking so long to reply.

This is the part of my code used to acquire data:

If you were talking about screen capture of a scope or something, I don't have it right away, but I am gonna try and see if I can make it work.

edit: I managed to get the scope working. Pictures are in the next page.

Hello again Brian,

I managed to get the scope working. Unfortunately since I don't have any kind of connector, I am using the probes (only 2 channels available) in order to get the signals.

Yellow line - DOUT

Green line - SCLK

The following image shows a whole 216 bits aquisition

The following image shows the very beggining:

And here is my DRDY, working as 250 SPS (as set in the register)

And here is my DRDY when actually converting (green line is SCLK)

Hey Brian,

The first two pictures I posted on this page show the individual bits coming from the device (DOUT). Isn't this what you were hoping to see?

If not I'm sorry, can you explain in other words what you want?
I can capture right away (scope and arduino serial monitor if needed).

Here is the individual bits of each channel in arduino (input short mode):

First picture: RAW BITS

Second picture: BITS converted FROM 2's complement.

Hello Brian!

Ah I see...

So, I checked and the bits coming from DOUT is actually the same I can read from Arduino serial monitor.

Conversion start: (status)

The very next 24 bits (channel 1)

And Arduino serial monitor

As you can see, Status and Channel 1 are correct (I didn't check the other channels).

Do you think I should check all 216 bits as well just to make sure?

Hello Brian,

So here is my schematic:

I may have good news though: After checking the connections, I noticed that VREFN wasn't grounded.

After grounding it, things looked better.

Now I am able to measure the input voltage with more accuracy. The input short mode looks more stable, also in normal input mode grounded inputs

gives me a very low voltage (0.000... V). And when I try to measure a 2.5V battery I have I get a 3.18V (at least it is stable).

As you can see now things are looking better.

In the following screenshot ALL channels EXCEPT CH6 are in input short mode. CH6 is in normal input mode, with a 2.5V ddp between IN6N and IN6P.

So, as I said, things are much better now. I still have this offset problem though (which may be caused by wrong calculation when converting bits to voltage).

And also CH8 is always "stuck" at this 4.49V even though it is in input short.

This is how I convert from decimal to voltage values:

channel[i]*(4.5/(16777215)    ----> I am dividing by 2^24 - 1

ps: channel[i] is already converted from 2's complement.

Thanks

Something I noticed: Using a power supply, I noticed that my if increase the input voltage, my output voltage displayed in arduino actually decreases.

Where is what I'm trying to say.

I am not sure about this but I changed my conversion formula to:

(4.5 - (channel[6]*(4.5/(16777215) ) ) )*2

It kinda works now but I did this without actually thinking, that's why I will double check it

Hey Brian!

Thanks for helping me, I figured out the problem in channel eight.

I wasn't including him in anything because in my iteration loop I stopped at channel 7. Very silly mistake!

Btw, I tried to do what you said with negative values and it isn't working. I got the logic of it and the steps are simple, but I tried something else

and it worked right away so I'll take a second look at the way you said in another time.

I found this in the internet:

 You can do this by 'XORing' the data with 0x800000. Full scale positive is 0x7FFFFF, XORing that with 0x800000 would give you 0xFFFFFF. Full scale negative is 0x800000, which yields 0x000000 when XORed.

Would it accomplish the same as you said?

I tried this method and noticed that my voltage values were ranging from 4.5 to 9v. Actually from 0 to 9v if negative was included.

So, I just subtracted 4.5 of my signal in order to "offset it back" to 0 again.

Now I get -4.5 to +4.5 scale.

Any errors in all this?

Hey Brian,

So the problem about the offset was what you said: the leads were floating with respect to the board supplies.

I grounded the battery (negative lead) to the negative of the power supply and that worked.

I'm using a battery (it's 2,2V now) and I'm getting the following:

As you can see there are some spikes. I've implemented a moving average to "cut" them out and it works perfectly.

But I am not sure the cause of it...

So I tried to implement a passive 60Hz filter but it didn't work (I gave it a shot even though it doesn't seem to be a 60 Hz power line noise problem).

Should I worry about them or they are just a result of my whole protoboard circuit?

Another question, if my PGA isn't working properly (high or low voltages, doesn't matter) then I should take a better look into the common-mode range?

Here are some tests (I used 2 power supply - one grounding INN and the other actually giving voltage to INN and INP).

GAIN - VOLTAGE

1 - 2.25 (correct)
2 - 3.03
4 - 3.64
6 - 3.90
8 - 4.05
12 - 4.21
24 - 4.37

1 - 0.107 (correct)
2 - 0.16
4 - 0.19