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.

TIDA-00796: Interfacing Resolver with TMDSDOCK28379D using TIDA-00796/BOOSTXL-POSMGR

Part Number: TIDA-00796

Hello,

I am working with TMDSDOCK28379D. I am planning on using resolver from LTN (RE-21-1-A-05). I was looking offerings from TI which can interface with the resolver and provide sine/cosine outputs. I have found following two solutions.

(1) BOOSTXL-POSMGR

(2) TIDA-00796.

I am seeking some guidance on the following doubts.

(1) Which one of these solutions is a better fit for resolver+TMDSDOCK28379D?

(2) As far as my research TIDA-00796 would be a better fit. Am I concluding this right?

(3) If I plan on using TIDA-00796, can it provide excitation voltage of 7Vrms at 7 kHz (TIDA-00796 Spec is 10-20 kHz)? As the resolver expects excitation between 5-7 kHz. Please see attached figure.

(4) Would I be able to interface sine and cosine outputs from TIDA-00796, directly to TMDSDOCK28379D? i.e. would they be in the range of 0-3.3V?

(5) Are there any other offerings from TI that can work with resolvers?

I would greatly appreciate any help.

Regards,

Nakul Shah

  • Dear Nakul, 

    thank you for the interest in the reference design. The PGA411-Q1 is no longer recommended as a good solution for resolver to digital conversion due to performance issues related to motor noise immunity. So we strongly recommend you consider alternate solution based on discrete amplifiers combined with a C2000 digital signal controller as outlined in TIDA-01527 and TIDM-02009.

    Still, answering your questions:

    (1) Which one of these solutions is a better fit for resolver+TMDSDOCK28379D?

    TIDA-01527

    (2) As far as my research TIDA-00796 would be a better fit. Am I concluding this right?

    * See the answer above

    (3) If I plan on using TIDA-00796, can it provide excitation voltage of 7Vrms at 7 kHz (TIDA-00796 Spec is 10-20 kHz)? As the resolver expects excitation between 5-7 kHz. 

    * The PGA411-Q1 unfortunately does not support the excitation frequency below 10kHz. The excitation frequency of 7kHz is not possible with the PGA 411-Q1

    (4) Would I be able to interface sine and cosine outputs from TIDA-00796, directly to TMDSDOCK28379D? i.e. would they be in the range of 0-3.3V?

    * Yes, this is possible by monitoring OCOS and OSIN signals.

    (5) Are there any other offerings from TI that can work with resolvers?

    * We suggest using the C2000 MCU together with the discrete resolver front-end as shown in TIDA-01527. For the excitation amplifier we highly suggest the ALM2402F-Q1 or ALM2403-Q1 device.

    Please have a look at 

    https://www.ti.com/lit/an/sboa504/sboa504.pdf

    Best regards, Jiri Panacek

  • Jiri,

    Thank you very much for guiding me through this.

    Just a follow up question, you have mentioned that we should use ALM2402F-Q1 or ALM2403-Q1 device for excitation amplifier. Does TIDA-01527 comes with either one of this?

    I use Matlab to generate code for my F28379D. Is it possible to generate SPWM on a PWM port for exciter without sinusoidal reference (without SW dependency)? For example, usually to generate sinusoidal PWM, I use sample based sine wave block as input to PWM block as shown in the picture attached. I am not sure if I can generate a sample base sine for the input of PWM block for higher frequencies. So is there a way to enable SPWM inside the PWM block so it doesn't require an input?

  • Hi Nakul, I am sorry it took me long to reply. The TIDA-01527 does not come with the ALM2402F or ALM2403-Q1. Nevertheless, I created an internal version with the IC used. Please find the circuit diagram in the attachment. 

    For the sinewave generation (modulation of the PWM) you can use a DMA. This is actually what we use in TIDM-02009. You need to generate a lookup table with sinewave coefficients and then the DMA automatically runs the update in the loop. Does this answer your question?

  • Jiri,

    I have sent the boards out to fabrication. Would the existing design wouldn't work? (Design files available for TIDA-01527)

    From the fabrication vendor I got some questions back and was hoping if you could help me with those.

    (1) The Gerber files mentions that there are some VIPs but it doesn't state which ones. Is it possible to just convert all VIPs in normal vias?

    (2) DMC564040R (Q1Aand Q1B) are not available to order anywhere. Is there an replacement part I can use?

    (3) Q1B and Q1A are using to control analog crossbar which is experimental. Is it possible to use the TIDA-01527 without crossbars thus eliminating the need of Q1B and Q1A?

    -Nakul

  • Hi Nakul,

    Don't worry, the design will work. 

    ad 1 - Do you mean by VIP "via in pad" ? I typically don't do it but I did this time.... I manufactured it always with just standard vias option (i believe 28/12mil)

    ad2) I believe PUMH9 will do the job and comes in the same package. 

    ad3) Definitely yes, just populate the 0-ohm resistors to bypass it. nevertheless, I suggest assembling it. It is the key element of this reference design (the scattered signal processing method).

    Jiri

  • Jiri,

    I would appreciate it if you could confirm which PUMH9 part would work.

    -Nakul

  • Hello Nakul, I've just saw your message. Yes, this PUMH9 shall work. I checked the resistance, the pinout and the package. It is identical to the DMC56404. 

  • Thank you Jiri. I just checked and minimum buying quantity for that PUMH9 is 12500. 

    Any chance https://www.digikey.com/en/products/detail/nexperia-usa-inc/PUMH9-115/1157884 this PUMH would work?

    -Nakul

  • Jiri,

    Thank you for your support so far. I have fabricated the boards with replaced components and was able to get a 5kHz sinewave as output. I am using F28379D control card with switching frequency of 312.5kHz. I have programmed the controller with MATLAB. Although at the output I am seeing some oscillations in the sine wave. Please see attached picture. Can you please help me mitigate that?