PGA411-Q1: Evaluation board support

Thank you very much for your prompt response. I greatly appreciate that.
Is it possible to feed the resolver reference signal from another source but have the feedback signals on PGA411-Q1?

Thanks a lot Clancy. Much appreciated your prompt response.

We are planning to use the device for position sensing for production scale.
We are planning to talk to PGA411-Q1 via SPI. I am wondering at what instant of time would be SPI registers be frozen?
In detail, I am trying to read the position value from DEV_STAT5. At what instant of time will this value be frozen to be sent out. Will that be when we drop the NCS pin from high to low or at any other instant.

Thanks for your support.

-Ram

Thanks a lot Clancy!
The external reference signals is of 9.78kHz. Is it ok to input this to PGA411-Q1 ?

Hi Clancy

Thank you very much for your support.

The PGA411-Q1 EVM is in fault mode and I am not able to get it our of the fault mode. When I measure 3.3V test pin, its measuring about 1.5V with 30Hz freq. I had the exciter sending a 10kHz ref wave form to our resolver but I guess the parameters on PGA411-Q1 got messed up.
Would it be possible to reset the chip to factory defaults. The angle I am reading from the chip makes no sense and its flipping between 0 and 0x3FF.

The fact that you are measuring 1.5V on the 3.3V rail is interesting. I have a few debugging questions for you:

- How much current is being drawn from your power supply?
- What are you measuring at the VCC test point (it should be 5V)?
- Do you still have your MCU connected to the PGA411 EVM?
- Is the exciter signal being generated by the PGA411 in this state?

Basically, let's make sure that everything is powering up correctly and then we can look at the register settings. No matter what the PGA411 is programmed to, you should see correct VCC, VIO, and VDD voltages from the EVM.

How much current is being drawn from your power supply?
0.18A.
- What are you measuring at the VCC test point (it should be 5V)?
I measure 5V.
- Do you still have your MCU connected to the PGA411 EVM?
No
- Is the exciter signal being generated by the PGA411 in this state?
No.

I got help from one of the TIs tech support and they are shipping me new PGA411 chips.

Good, that will be helpful.

Are you still seeing the lower voltage on the 3.3V rail? If 3.3V is not supplied to VIO, that would explain why you are unable to read registers. Have you tried changing J3 to select 5V for VIO? (connect pins 5 and 6 of J4)

-Clancy

Yes I am still seeing low voltage on 3.3V rail as before.
If I change the J3 jumper to 5V, would it affect MSP because it runs on 3.3V right?

Thank you very much!

Thanks Clancy
I have the EVM hooked up to our inverter that controls the motor. When I start switching the inverter, the SPI lines are getting corrupted due to switching noise. Do you have any noise mitigation techniques ?

Hopefully this shouldn't be an issue once you create a new PCB since you won't have the SPI lines connected by wires anymore. In the meantime, here are a few ideas:

- Use short wires to connect the boards.
- Improvise shielding for the wires. (SPI wires and wires to the motor)
- Route the motor switching signals away from the SPI lines.
- Make sure that your ground connections are solid. Try connecting both the PGA411 and the MCU board to Earth grounds.
- Check that the power supplies to both the PGA411-Q1 and the MCU board are stable and not corrupted by the switching noise.

Hi Clancy

I tried as suggested but I still see the issue.
I have also noticed that the reference signal generated from the PGA chip is getting corrupted once the inverter starts switching. I disconnected the SPI lines from the MCU board and I also disconnected the resolver signals going to the motor. The EVM board signals are all disconnected except the 5V power supply. Even with this, the reference signals measured on the scope is clean when the inverter is not switching but soon when the inverter turns "ON", the reference signal is getting corrupted. Our switching freq is at 10 kHz.

As I mentioned, I disconnected the SPI lines from the MCU board. I do not think the noise interference is on the SPI lines. The issue is with the reference signal thats generated by the PGA thats getting affected.

So the exciter signal from the PGA411 is affected even when no resolver sensor is connected, correct? We need to identify the coupling path for the noise. It must be coming from ground, the power rail, or the radiated emissions from the inverter are being picked up by the EVM. You should be able to see the noise on VCC on an o-scope if it is coming from the power supply.

Note that I have seen the switching inverter noise show up on the exciter signals before (when everything is connected). I haven't seen this cause issues with the signal path, especially with the filtering on the sin and cos signals, although filtering will not be able to do too much if the switching frequency is at 10kHz since it is so close to the exciter frequency.

This switching noise showing up on the SPI signals is more of a problem however since that can corrupt your SPI read/writes.

If you still need support for this noise issue, then please start a new thread. Thanks!