• State Resolved
  • Locked Locked
  • Replies 16 replies
  • Subscribers 31 subscribers
  • Views 693 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.

PGA411-Q1 sudden rotation speed spike

Alexander Bil
Prodigy 90 points
Other Parts Discussed in Thread: PGA411-Q1

Hello everyone,

we are using a PGA411-Q1 in our project. We are using the the SPI-Interface of the "resolver chip" to read the rotation speed as well as the serial interface to read the resolver position and calculate the rotation speed ourselves. Both variants work fine and we get the same result.

During some testing in a vehicle we have seen a sudden, implausible rise of the rotation speed with both ways of determining the rotation speed. The spike lasts a couple of hundrets of ms and we see accelerations of about 4000rpm during 50ms.

The state of the chip is always in normal. No fault is reported via SPI nor on the FAULT PIN.

What could be a cause of such a behaviour. Is this issue maybe already known, is there a fix for this?

Thanks in advance!

  • 0
    TI__Mastermind 40295 points

    Alexander,


    Many of our customers have struggled to get to production with this part, so we recommend a discrete solution over the PGA411-Q1.
    An example of a discrete solution can be found at https://www.ti.com/tool/TIDA-01527 .

    Having said that, is there a pattern to the estimated speed chnage, or does it appear randomly?

    You should see some pulses on the OUTA, OUTB, and OUTZ pins that represent incremental encoding. 
    Do they have constant widths, or do they vary when the part indicates the increase in speed?

    regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello John,

    we have this controller already in use in quite a few devices. Untill now we were not seeing any unreasonable issues at our resolver interface.

    These spikes seem random, one pattern is visible. They appear at low speeds or at standstill. Or maybe they are jsut unnoted at higher speeds as the influence is much less.

    We do not have the OUTZ, OUTA and OUTB connected on our PCB so I am not able to provide any feedback on this.

    Are there any known conditions which could lead to a sudden increase of speed?

  • 0 in reply to Alexander Bil
    TI__Mastermind 40295 points

    Alexander,

    Thank you for the reply.
    How many units are showing this problem?
    What device outputs are you able to look at?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello,

    we have seen this behaviour in two units now. We have several hundrets running without this issue.

    We are looking at the ORD0-11 to get the angle and do the caluclation of the rotation speed on our own, the parity bit is always taken into account to validate the transmission of the angle. And we are reading the angle and rotation speed via SPI. And we also do status checks of the DEV_STAT Registers. Additionally we are also polling the FAULT pin to check for any faults.

  • 0 in reply to Alexander Bil
    TI__Mastermind 40295 points

    Alexander.

    Thank you for the update.

    Are the units interfacing with a resolver during testing?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello,

    yes. On the IZ1,IZ3 the Cos, on the IZ2, IZ4 the SIN, on the IE1, IE2, OE1 and OE2 the Exciter signals are connected. The resolver kind does not vary from normal to the noticeable cases.

  • 0 in reply to Alexander Bil
    TI__Expert 7495 points

    Hi Alexander,

     I wanted to let you know that John is out of the office now.  He should be able to follow up with you next week when he is back in the office.

    Regards,

    Mekre

  • 0 in reply to Mekre Mesganaw
    TI__Mastermind 40295 points

    Alexander,

    Would it be possible to swap resolvers between units that work, and units that don't, to see if the failure follows the resolvers?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hi,

    it is not that easy to swap the resolvers as they are mounted on the motor shaft. They are serial parts. We havent had a problem with them. In operation at higher speeds the signals are not noticeable.

    We are currently investigating some other possibilities for the issue. We think of some mechanical stress / vibrations on the shaft. I can let you know about the outcome later on. To be honest we do not rely believe this being an issue because the accelarations and total change in angle would indicate the values that the shaft is not able to withstand and this would be definetly noted.

    So another consideration has to be, that this values are not real. Are there any known cases which lead to excessive angle changes from the resolver chip?

    For example noise on the resolver signals, undervoltages, internal chip failures which lead to a distortion of the sample or calculation algorithm inside the chip?

    Thank you for your support!

  • 0 in reply to Alexander Bil
    TI__Mastermind 40295 points

    Alexander,

    One last question before I dig further into this further: does the observed "spike" have any dependency on the actual motor speed?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello John,

    in the 2 cases we have seen right now, there does not seem to be any correlation, they both appeared during slow speeds or at standstill. We have had one spike to positive rotation speeds up to 4000rpm and a negative spike to values exceeding -8000rpm, values beyond -8000rpm are not transmitted anymore, so I am not ble to say what the actual peak value was.

    Hope this helps.

  • 0 in reply to Alexander Bil
    TI__Mastermind 40295 points

    Thanks Alexander.

    I will dig into the archives to see if this has come up before, and will update the thread by COB tomorrow.

    John

  • 0 in reply to John Miller - Sensing
    TI__Mastermind 40295 points

    Alexander,

    Are you using auto-mode phase correction in your application?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello John,

    yes we are using the auto-mode phase correction, but we are using it with the Open Fault Detection Threshold Levels (Setting Bit1 = 1, Bit0 = 0).

  • +1 in reply to Alexander Bil
    TI__Mastermind 40295 points

    Alexander, 

    It has been found the device can report changing angles when the motor is stationary.

    The conditions that can cause this behavior:

    1. AUTOPHASE_CFG set to 01 or 10 instead of 00
    2. OSHORTL and OSHORTH set too large such that OCOS and OSIN never cross the thresholds
    3. FAULTRES held low (ignoring faults) while improper signals are input to the PGA411
      • No sensor connected
      • Mutual short
    4. OSIN/OCOS or IE1/IE2 signals are out of range
      • OSIN/OCOS should be around 1.5Vpp max
      • IE1 and IE2 should be within 0.5V to 4.5V
    5. Phase delay between IEx and IZx inputs is too low.
      • Phase delay around 0 degrees causes issues for the PGA411-Q1 auto-phase algorithm.
      • To remedy this, implement a filter with at least 20 degrees of phase delay.

    Also, the device specs for Exciter to SIN & COS phase adjustment range were revised to improve the integrity of the tracking loop, which can in turn impact the angle tracking & reporting.
    The previous limits for auto mode were -89.2° to 89.2°, and the updated limits are  -89.2° to -20° and  20° to 89.2°.
    These limits apply when the exciter signal is between 10kHz and 20kHz. 

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 90 points

    Hello John,

    thanks for the continous support. We will check on these points.