• State Resolved
  • Locked Locked
  • Replies 45 replies
  • Answers 3 answers
  • Subscribers 33 subscribers
  • Views 4999 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LDC1612EVM: Can we make changes to EVM firmware for different measurements?

Ali Nekourouh
Intellectual 345 points
Part Number: LDC1612EVM
Other Parts Discussed in Thread: LDC1612

Hi,

Currently the LDC1612EVM takes frequencies in the range of 1MHz to 3.3MHz (deglitch filter). What if we want to change the frequency range to 10kHz to 100kHz range, do we need to do this in the firmware only? or hardware as well? Any hints are appreciated.
Also, I will follow section 2.9 within User's guide for updating the EVM firmware (unless you think otherwise).
Thanks.

  • 0
    TI__Genius 11015 points
    Hello,

    We have received your question & will get back with you in the next few days.
  • 0
    TI__Expert 6625 points
    Hi Ali,
    As you can find from section 8.1.7 from the datasheet www.ti.com/.../ldc1612.pdf. The deglitch filter only has four settings, b001(1MHz), b100(3.3MHz), b10(10MHz)1, b011(33MHz). If you would like to choose the range from 10kHz to 100kHz, I would set MUX_CONFIG.DEGLITCH = b001. Again, always make sure that this bandwidth is higher than the sensor frequency. If you really want to narrow down from 10kHz to 100kHz, you can always use a bandpass filter to achieve this. But again make sure that the sensor frequency is below this range. I hope this helps!

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    Thank you for your reply. In regards to setting MUX_CONFIG.DEGLITCH=b001, do you think that would be enough by itself or do I need to consider your second option which is to use a bandpass filter? Let's say I use the bandpass filter, do I still need to set the MUX_CONFIG.DEGLITCH to 1MHz setting? I am assuming the bandpass filter is to be implemented in LC tank section of the circuit, correct? Thank you.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    For you case, I would just change the MUX_CONFIG.DEGLITCH = b001 (1MHz). You won't need a bandpass filter. I hope this helps!

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    Can you please verify that this to be done in the firmware (source code) or on the application (GUI)? Looking at the GUI, I was getting the impression that perhaps it is on GUI side... Thank you.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    Intellectual 345 points
    Hi Bala,
    Another question: Is the firmware done in Code Composer Studio or IAR? The GUI seems to be done in Python, an open source one would do? Thank you.
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    You should be able to set the deglitch setting in the "Register Map" page of the GUI. For details on the firmware and GUI, please accept my friend request, I will share that information in private.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    Thanks. I did accept the request.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    I just sent you the requested information.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    I sent you a message yesterday after I tried the change.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    Intellectual 345 points
    Hi Bala,
    While I am still waiting for your answer on the deglitch filter issue, I am thinking if we bypass the UI overall and measure the sensor frequency on the board, are IN0A and IN0B targets we should measure? Thanks.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    Intellectual 345 points

    Bala,
    Sorry with my messages, I am finding more as I go along (I feel it's worth sharing)...While reading the datasheet, I noted on page 50 under "8.2.4 Detailed Design Procedure, 1 b) The design constraint for ƒREF0 is > 4 × ƒSENSOR." and the 40MHz reference frequency on the board satisfies the condition for EVM. If we want to have the sensor frequency's range within 10kHz-100kHz, then we need to have at least a reference frequency of 400kHz (taking the maximum in this case), correct? Reading through the EVM Guide, section 2.7.2 (page 25), it states "The EVM includes a 40-MHz oscillator on-board but an external off-board signal can be used.", so how can we use the external off-board signal?
    Also, when setting the register MUX_CONFIG to 0x0209, I read 2.19MHz for sensor frequency with "input deglitch filter incorrect" being lit red, whereas when I select 3.3MHz as input deglitch filter, it's not red anymore... which explains why we need to select the one that exceeds the oscillation frequency.
    One thing I thought it's worth mentioning is that when I brought External Oscillator value (under Reference Clock Source) in UI to lower than 10MHz, the red button is not shown anymore, the frequency in UI changes (in decimals), however when I measure on the board still reading the same 2.19MHz, which is due to the external clock (40MHz for LDC1612) on the board.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    Intellectual 345 points
    Hi Bala,
    I think I figured how to replace the reference clock on the EVM, if you can verify that would be great! Looking at the schematics, R22 has to removed and placed on R14 and R11 has to be populated. LDC1612 datasheet has reference clock listed between 2MHz and 40MHz, for our case, does it make sense to go with 4MHz? Still waiting for answers to my other questions. Thanks.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    I will verify tomorrow if that's the right way to replace the reference clock. About the reference clock, you can pick any frequency within the range, but note that higher the frequency more resolution you can get on your results. For your case the minimum requirement is 4 x fsensor = 400kHz. Since the minimum fref is 2MHz, you need to at least go with this clock frequency. But, you can always go higher for more resolution. I've also replied to you on private message, if you've any other unanswered questions, please let me know.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    Thanks for your answers.
    Also, this question I had "Reading through the EVM Guide, section 2.7.2 (page 25), it states "The EVM includes a 40-MHz oscillator on-board but an external off-board signal can be used.", so how can we use the external off-board signal?", can we use a function generator instead of the reference clock (in your previous message)? Thanks.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points

    Hi Ali, 

               Sorry about the delayed response. Yes, you can use a function generator instead of the reference clock. As you can see from figure below, you can apply the clock to R14 and you need R11 of 50 ohms and you need to short R22. I hope this helps!

    Best Regards,

    Bala Ravi

  • 0 in reply to Bala Ravi
    Intellectual 345 points
    Hi Bala,
    Ok, tried it with the function generator and made those changes to the board... however the input "deglitch filter incorrect" is lid red again. I have tried different reference frequencies between 2MHz-20MHz on the function generator and the result is the same. Just to make sure, "External Oscillator" under "Reference Clock Source"on the UI has to match the reference clock I choose, correct? The reason I ask is cuz when I try let's say 10MHz on the function generator and then choose 4MHz as reference clock source on UI, that deglitch filter incorrect indicator on the UI is turned off. Also, I noted on the schematics it says to populate R14 instead of R22 (R14 or R22), that means R22 is to be open... but you had indicated in your last post to short R22, can you please verify on this? Thanks.

    Regards,
    Ali
  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Can you please share a picture of the "Deglitch filter incorrect" is lit red? Yes, your external oscillator should match the ref clock on your GUI. Sorry about the confusion earlier, R22 must be open when using an external clock.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Please see below a screenshot of UI while measuring at 10MHz reference clock, somehow it doesn't display the frequency now... earlier it was showing the "deglitch filter incorrect" as stated in my previous mUI_screenshot.docxessage.

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Did you choose the smallest possible frequency (for deglitch filter) that is greater than your sensor frequency? Would it be possible to share a screenshot of the error screen?

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Answer to your 1st question is yes, I did. Please see attachment that contains 2 screenshots for 3MHz and 5MHz reference clock (for some reason "paste from word" didn't work).UI_screenshot1.docx

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Does this error also happen when the deglitch frequency is set to 3.3MHz?

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    No, the "deglitch filter" is not red for 3.3MHz, but then "Saturated" is turned red for 3.3MHz, 10MHz and 33MHz. When I try the same for 5MHz reference clock, those are not indicating red.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Allow me some time to try this lab. I will get back to you either by today or tomorrow.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    TI__Expert 6625 points
    Hi Ali,
    I just tried this in lab. The reason you're seeing this error is because you're trying to set the deglitch filter bandwidth lower than your sensor frequency. Either lower your sensor frequency or up your deglitch filter settings. Also make sure your Fref is at least 4x Sensor Frequency.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    I don't understand! Please refer to my messages from October 11th and you will see that we have been discussing this over... Let's look at numbers here: So my sensor frequency is to be in the range of 10kHz-100kHz and based on that and the Fref > 4 x sensor frequency, I have made the change to the EVM and replaced those components and provided external clock of 2MHz or higher (according to LDC1612 datasheet). So if I choose 1MHz deglitch filter in UI for measurement, how is the deglitch filter bandwidth lower than sensor frequency? again, having sensor frequency in the range of 10kHz-100kHz, Fref >2MHz. 1MHz deglitch filter setting is the highest (and closest) to my sensor frequency, not 3.3MHz in this case... correct? Thanks.

    Regards,

    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    What are your inductor and capacitor values? As you can see from the column to your right "Calculated Sensor Data", your sensor frequency is 2.192MHz for 5MHz clock signal and 1.889MHz for 3MHz clock signal. Can you please verify that under Reference Clock Source for "External Oscillator" you entered 5MHz and 3MHz? This should exactly match the frequency from your function generator. Finally for your application, is it possible to work at a higher bandwidth, greater than 150kHz? At lower frequency the skin depth is larger and your target needs to be at least 2x or 3x skin depth. If you haven't already done so, please download our "Inductive Sensing Calculator" at www.ti.com/.../slyc137. You can calculate all your parameters using this tool.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Please see answers to your questions below:

    - The values for inductor and capacitors are the same as what you note on the schematics of LDC1612 evaluation board, for C0 channel there are 2 330pF caps in parallel. I am not using anything more than those components.
    - Yes, I did enter those frequencies (3MHz and 5MHz) for external oscillator values in UI.
    - The problem with choosing higher bandwidth from our findings is that the stray capacitance for our application would be too high for LC circuit, thus would would make it unstable. Although we would be open to new suggestions if you think it makes sense to go for higher bandwidth.

    My question is do you think this LDC1612 EVM is suited for running in 10kHz-100kHz range for sensor frequency? Thanks.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Yes, LDC1612 is well suited for a sensor frequency between 10kHz -100 kHz. What's your target size? As I mentioned before, the target needs to be at least 2x or 3x the skin depth. Can you also please provide a picture of your setup?

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    I am only measuring the frequency of Ch0 without connecting anything to it (no target beyond EVM) for now. I should be able to measure for low frequency with the settings we talked about, correct? My setup consists of the EVM and applying external clock (>2MHz) at CLKIN on the board. You said LDC1612 is well suited for a sensor frequency between 10kHz and 100kHz, which is what the datasheet says... but how about the eval board LDC1612EVM that we are working with? Because of this experiment it seems that something is missing or... as we are not able to measure it.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    First off, the "input deglitch filter" incorrect light is calculated by GUI, not the IC. If the calculated frequency is higher than the deglitch sensor, then the red light is flagged. You can try to measure one of the INx pins via an oscilloscope, to see what frequency is seen there. That way you can see if your sensor frequency is actually higher than your deglitch settings. Lastly, you need to change your LC value if you want to test a lower frequency sensor, because the device will have to drive to that frequency.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Thanks for your reply. I will measure the frequency on the board and let you know. To make sure the LC value you are referring to, we are talking of f = 1/ [(2pi) x sq. root (LC)] formula, correct? Also, one the eval board there are 2 330pF caps in parallel for C0, I think it's using inductance within the PCB... how can we change this to fit our criteria? Thanks.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Yes, that formula is correct. You can replace C24 and change the capacitance to fit your needs. You can also detach the sensor and connect your custom sensor to CH0. This way you can change the frequency of operation.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Based on that formula and the using the value of inductance (L) in UI, for range of frequencies of 10kHz to 100kHz, capacitance for C24 should be in the range of 0.33uF to 30uF. I chose 1uF as I prefer to be closer to 100kHz range. As I tried it with the new capacitance value and setting the external oscillator to 3MHz, the sensor frequency obtained (in UI) shows 3MHz and "Saturated" under "Measured Sensor Data" is lid. If 1MHz is selected for "Input deglitch filter", the "Input deglitch filter incorrect" is lid again... otherwise only "Saturated" is lid for other selections. What do you think is wrong here?

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    Intellectual 345 points

    Hi Bala,

    Haven't heard from you yet... it's been about a week. Any update? Thanks.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    What values of L and C did you choose? Is your resonant frequency below the deglitch filter setting? When you click on the data streaming tab in the GUI, and click start, does the output show reasonable values?

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    Based on the value in the GUI given for inductance, that's what I used. The value was 8.48uH. Using the formula we talked about with L and the frequency of 10kHz, C is about 30uF. Same with 100kHz and about 0.3uF is obtained. That is why I indicated in my last message that I chose 1uF. Yes, the resonant frequency is below the deglitch filter setting. I will try the data streaming tab and get back to you. Thanks.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Sorry about the delayed response. For the parallel inductance in the GUI the units are set to pF, so for 1uF you need to enter 1000000. Can you check using a scope on one of the INx pins what frequency the device is operating at? This is just to cross check calculated frequency to the actual frequency.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    I entered 1000000 for capacitance value in GUI. Using the scope, I am seeing 3MHz for frequency (all the time) at INx pin, however on the GUI it changes at times... Thanks.

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    Intellectual 345 points

    Hi Bala,

    Any updates? I have been waiting for sometime...

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Can you please share a screen shot of your GUI configurations page and register settings page? Also can you check the data streaming page and let me know if the measured data is reasonable.

    Best Regards,
    Bala Ravi
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala

    Attached herein please find the document that has screenshots of configurations page, register settings and data streaming for the measurement I took. With all the changes e did, I as expecting to see a frequency much lower than the one displayed... based on our discussions, e should see a frequency in the range of 10kHz to 100kHz. WUI_screenshot2docx.docxhat do you think is missing or wrong? Thanks.

  • 0 in reply to Ali Nekourouh
    Intellectual 345 points

    Hi Bala,

    Were you able to see the file I attached to my last message? Any updates?

    Regards,
    Ali

  • 0 in reply to Ali Nekourouh
    TI__Expert 6625 points
    Hi Ali,
    Yes, I did get the attachment. Thanks for sharing that information! The reason why you're having the deglitch error is because of your fref(external Oscillator). The digitized output for each channel is proportional to the ratio fsensor / fref. In your case, we are changing both these parameters hence we are not getting desired results. As you've already prepped the board for the desired fsensor range (about 55KHz), using L = 8.28uH and C = 1uF, you must use the default external clock frequency ~ 40MHz. Also for best performance you need to use the max permitted reference frequency (reference clock), this is stated in Section 7.3.4 of the datasheet. If resolution is not a major concern for your application, you can just use the internal clock with a typical clock frequency ~ 43.4MHz. As you can read from Section 8.1.6, external oscillator is only needed for precision applications. Finally, if you're worried about resolution, please refer to the following application note to optimize L-measurement: www.ti.com/.../snoa945.pdf.

    In conclusion, your issue should be resolved if you change the clock frequency to 40MHz. I will go ahead and close this thread, if you've any additional questions please start a new thread.

    Best Regards,
    Bala Ravi

    Please click "This resolved my issue" button if this post answers your question
  • 0 in reply to Bala Ravi
    Intellectual 345 points

    Hi Bala,

    I was finally able to measure lower frequencies with the same 3MHz clock frequency, which is good news. I will also try with 40MHz as you indicated to see if anything improves. I wanted to let you know about this. Thanks for your help.

    Regards,
    Ali