• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 2469 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1292RECG-FE: RLD and How to use it?

Frank Roberts
Intellectual 355 points
Part Number: ADS1292RECG-FE
Other Parts Discussed in Thread: ADS1292, ADS1292R, ADS1298

Hello,

  I have based a PCB design from the ADS1292R-FE Eval board schematic.  I would like to understand how to go about correctly using RLD in being able to derive an ECG waveform.  I get it that RLD can be used to reduce common mode voltage/interference.  What I am not fully understanding is how to actually make use of the RLD electrode/signal in deriving say a LEAD 1 (RA-LA) output signal.  For LEAD 1 on the ADS1292R-FE Eval from everything I can see if should simply be the output on CH2 of the ADC. 

So now my questions...

1) To make use of the RLD to reduce common mode voltages in the readings from the ADC is the idea to measure the RLD voltage on each channel and then subtract this value from the ADC channel output? 

2) The explanation of the RLD functionality in the ADS1292 Explains the Mux functionality of the RLD but does not explain how to actually make use of the RLD drive.  The description on page 50 of the DS shows RLD signal generated from CH1 and then fed into CH2 Negative?  It does not explain why anyone would do this.  It would be very useful to have an example shown with the ADS1292R-FE Eval board.

I would really like to know how to make use of the RLD signal to provide better results with the ADS1292R with the assumption the hardware is configured as the ADS1292R-FE Eval board is with electrodes matched to ADC channels in particular ELA & ERA connected to IN2P and IN2N respectively.  So how does one effectively make use of the RLD measurement/signal?

3) For the case of the ADS1292R-FE Eval board and in reviewing the MSP430 firmware I could not find anywhere where the RLD signal was used in conjunction with an ADC Channel?  Maybe I missed something.

Regards,

Frank

  • 0
    TI__Mastermind 38745 points

    Hello Frank,

    Let me just begin by explaining the purpose of RLD. In a typical ECG system, you might have 3 or 4 electrodes. One of those electrodes is called "RL". If you read the definitions of the ECG leads, you'll notice this electrode is not actually used to form any leads. The RL electrode serves two purposes:

    • Bias the patient. If the electrodes are all DC coupled to the patient, there needs to be some way of biasing the patient such that the electrode common-mode voltage is within the rails of the AFE. The average RL voltage will typically be (vdd + vss)/2. This voltage is applied to the patient putting his/her dc bias right in the middle of the AFE's common-mode range.
    • Reject common-mode signals. There is so much common-mode interference in the environment mostly from 50/60Hz mains frequency junk (i.e. florescent lights, fans, motors, etc.). This stuff easily couples on to the patient capactively. From an electrical engineering perspective, this does not initially sound like a problem, because the AFE can have very good CMRR. However, the impedance of the connections between the electrodes and the patients can be vastly different and cannot be controlled easily. This causes massive impedance mismatch between the inputs. So bad, in fact, that it does not matter how much CMRR your AFE has since the CMRR of the electrodes is terrible. To get around this, you sense the electrode common-mode voltage using the AFE, and invert that signal and apply it to the patient via the RL electrode. This does a good job of getting rid of most of the interference.

    Now I can answer your questions:

    1. Simply apply the RLD output to the patient via the RL electrode via a patient protection resistor.
    2. Figure 52 actually has the relevant diagram winch shows how to close the loop between the inputs and the RLD amplifier.
    3. Hopefully you can see why this would be the case given the above information.

    Regards,

    Brian Pisani

  • 0 in reply to Brian Pisani
    Intellectual 355 points

    Hello Brian,

      Thank you for all the responses.  Your responses have cleared up a LOT as to my understanding of the ADS1292R, its eval board, derivation of LEAD 1 from the ADS1292R-FE Eval configuration, and RLD usage. 

    All that being said the PCB I have developed is having significant issues with being able to produce a relatively clean LEAD 1/ECG waveform.  As mentioned before I have essentially replicated the ADS1292R-FE Eval board schematic.  We know our firmware that we use to interface to my custom ADS1292 PCB is functional as we can attach to the ADS1292R-FE via wire jumpers and capture a relatively clean LEAD 1 waveform on CH2 with RA, LA, and RLD electrodes.  We have implemented a LPF that also has helped significantly reduce noise levels. 

    Would it be possible to have someone review my Gerbers?  Depending on what PCB toolset you guys have I could zip the entire PCB project and send to you.  I just don't want to post this directly to the forum. 

    Possible Culprits:

    1) 51 vs. 51K Electrode/Anti-alias resistor on front end of ERA, ELA inputs.  You indicated this is primarily an RC filter for anti-alias in which I would still think we should see something resembling a ECG waveform using the exact same firmware/ADS1292 Register settings we can use to interface to the ADS1292R-FE and get a relatively clean looking waveform.

    2) It somewhat smells of a Grounding issue potentially.  One side of my PCB I have a Ground Poor for the Digital Ground(GND) and the other I have a Analog Ground(AGND).  There is a separation gap between the 2 ground copper pours and these are connected through a 0ohm resistor.  You will see this on the schematic attached.

    a) The Digital Ground connects to the SPI/Digital interface side of the ADS1292 and then to a 2x5 1.27mm header that is used to connect to my main Sensor Fusion/BLE board.  This same connector interface has been used to connect to the ADS1292R-FE successfully.

    3)  Another difference between my PCB and the ADS1292R-FE is the initial Power Supply filtering.  This is referenced from  the Multi-Parameter BioSignal Monitor design. 

    Outside of those differences we are completely at a loss as to why we can get even what we would call a relatively clean signal out of our own custom ADS1292R board when it so closely resembles the Eval Board?

    Please let me know if either yourself or if you can point me to someone that I can have review the actual PCB/Gerber files.

    Regards,

    Frank

    5001.smSFM1-ECG SCH.PDF

  • 0 in reply to Frank Roberts
    TI__Mastermind 38745 points
    Hello Frank,

    I actually tend to think the layout may not be the culprit. As long as the device is connected functionally correct, I think you should be able to measure something decent. Rather than jumping into that, can you send me your full register settings? Then I can check to see if the device is configured in the best way. Wrong configuration tends to lead to a lot of problems problems, so I think that should be the first place we look.

    I also don't think the resistor value should be a major issue. That RC filter does provide anti-aliasing, but there is still another anti-alias filter on the chip, so the performance should not be terrible. Plus the band that allows for aliasing is pretty small since this is a delta-sigma ADC.

    Brian
  • 0 in reply to Brian Pisani
    Intellectual 355 points

    Hello Brian,

     Thanks again for all your assistance.  I have attached a Word Doc that I have captured the Screenshots of the ADS1292-FE register settings that we have used to get a good LEAD 1 waveform from the FE Eval board.  The only difference between the 2 is some LEAD OFF configurations.  You will also see a screenshot of the LEAD 1 waveform with a 6tap LPF enabled with the Fc set to 35Hz which is the same LPF filter configuration we use in our firmware.

    What doesn't make any sense and why it seems like there has to be some sort of issue with either a GND'ing is that we use the exact same firmware to connect to the ADS1292R-FE as we do to connect to our own custom hardware and we simply get what appears to be random noise on our own hardware where we get a well distinguishable ECG waveform on when connected to the FE Eval?  In fact I would have expected much more noise when connected to the FE-Eval board as I had to kluge together a 2x5 1.27mm connector and attach leads to the connector with 30 AWG wires that are soldered to all the test points on the FE-Eval board to communicate with it from our base Sensor Fusion module.  Whereas our own custom board has a clean connector/cable combination. 

    ADS1292_EvalBrd_Settings.docx

  • 0 in reply to Frank Roberts
    TI__Mastermind 38745 points
    Hey Frank,

    I'm aasuming that the second register map image in the file you sent is your register settings for your custom hardware where you are seeing the problems. Is this correct?

    If so, it should be fine. Have you tried measuring the internal test signal or the internal short? If you set the bits [1:0] of CONFIG2 and change bits [3:0] of CH2SET to 0101, you should be able to measure a 1 mV, 1Hz square wave. You can try this out with the EVM first to make sure the settings are right.

    If you can measure that with the EVM, but not with your hardware, there may be some issue. in any case could you take a picture of the data you collect and post it here?

    Brian
  • 0 in reply to Brian Pisani
    Intellectual 355 points

    Hi Brian,

      Again Thanks for all your help.  The 2 settings shown the only difference is whether some of the Lead Off detection is enabled and have been used successfully with the ADS1292R-FE.

    I should have mentioned earlier that we have successfully got the Test/Square wave out of the ADS1292R on our Custom Hardware.  This is one of the first tests we did once we got the custom hardware back from our PCB house.  We have done work also with the ADS1298 and used this test signal also on that so we have familiarity with this.

      The issues only arise once we put the ADC channels in Normal Electrode.  I have attached captured data from both the FE-EVAL board and our own custom board.  In each of the spreadsheets the data in column B & C is captured from CH2.  Column B is RAW data whereas Column C is the data after being fed through a 35Hz 6 tap IIR LPF Filter.  You will see we get a distinguishable ECG waveform from the FE-EVAL board whereas our board it just looks like random noise. 

    Some Notes:

    ECGAddon_xxxRun1 => Swapped what would be LA & RA electrodes to the input of the custom hardware.  I.E. - LA electrode on body connected to RA input on custom hardware.

    ECGAddon_xxxRun2 => Normal Electrode to input.  You will note there seems to be a significant Negative DC offset?  This is the reason I tried switching the electrodes.

    It almost seems to me that one of the following is going on...

    1) There is some sort of GND contention/issue.  I do have separate GND pours for AGND and DGND in which they are effectively connected through a 0ohm resistor as shown on the schematic.

    2) The Negative DC offset somewhat makes me think there is possibly something screwy with the RLD circuit/electrode drive?

    Do you have any other suggestions to test/try?  We are running up against some deadlines to get this working...

    Regards,

    Frank

    20170620_ADS1292R-FE_CH2, CH2 Filtered_Run1.xlsx20170620_ECGAddon_CH2, CH2 Filtered_Run1.xlsx20170620_ECGAddon_CH2, CH2 Filtered_Run2.xlsx

  • 0 in reply to Frank Roberts
    TI__Mastermind 38745 points
    Frank,

    It does not even look like the ECG signal is in there. Can you use a normal sine wave generator and measure a simple 1 Vp 10 Hz sine wave? If not, I recommend you probe with an oscilloscope at different places on the board until you find the issue. You can measure the source, you can measure after the RC filters, and you can even measure the PGA output at PGA2P and PGA2N. Remember that you will need to bias the signal generator output so I recommend connecting LA to the positive output, then tie LA and RL together and connect that to the signal generator ground.

    I would not read too much into the offset voltage just yet. It is common for ECG electrodes to impose up to a few hundred millivolts of offset between them.

    Regards,
    Brian
  • 0 in reply to Brian Pisani
    Intellectual 355 points

    20170626_smECG_CH2,LEAD1_Gain12_51_run5.xlsx20170626_smECG_CH2,LEAD1_Gain12_51K_run6.xlsxHello Brian,

     Well the ECG signal was in that waveform just couldn't make it out.  As I had suspected there was a GND'ing issue.  In closely reviewing the PCB files I noticed that several VDD/Power supply caps did not have a straight path to the L2 GND plane.  I should have put a couple of vias on the GND side as I had the ADS_DVDD side.  So I took one of my boards and soldered a 30 AWG wire across the tops of the caps as shown in the attached WORD Doc to another cap which had a via directly to L2 GND plane.  L1 had a solid GND pour seen in the solid Red but from other experience I have seen that long GND path returns can cause issues.  Immediately after putting this fix in we were able to start to see an ECG waveform see Run 5 attached.  I then made the change from the 51ohm to 51K ohm for the front-end anti-alias RC LPF and you can see the results in Run 6.  Run 6 produced as good if not better ECG waveform results as the ADS1292R-FE Eval board.  Amazing what a very minor change in the GND connection can produce.  

    Regards,

    Frank

    Ground Attach.docx

  • 0 in reply to Frank Roberts
    TI__Mastermind 38745 points
    Hello Frank,

    I'm glad that you got it working. I apologize for doubting your intuition originally.

    Brian