TMS570LS3137: I'm trying to test SPI4 on TMS570LS3137 devkit.

Hi IBLEE,

INBUM LEE said:

(K18)SPI4_CLK
(U1)SPI4_CS0
(V6)SPI4_SOMI
(W5)SPI4_SIMO

Yes, these pins are correct.

I will explain how to test SPI4 on devkit.

If you verify schematic, the pins are named as below:

NHET100 - SPI4CLK
NHET102 - SPI4SIMO
NHET103_W2FCPUENON - SPI4NCS[0]
NHET105 - SPI4SOMI

Note: There is a mistake in the name for SPI4CLK, it is named as SPI2CLK instead of SPI4CLK.

If you verify above pins, they were terminated to the J11 connector.

NHET100 - J11(38)
NHET102 - J11(40)
NHET103_W2FCPUENON - J11(39)
NHET105 - J11(41)

So, you can use above mentioned J11 connector pins to test SPI4.

Here you have to make sure one more final thing is that NHET103_W2FCPUENON this pin has multiple functionalities, and it is terminated through a Demux

As you can see, NHET103_W2FCPUENON is used as OHCI_SPEED1 and W2FC_PUENON_OHCI_SPEED1(NHET103).

In our case we need NHET103 to test SPI right, so in this case this USB_HOST1_ON selection line should be in OFF.

--
Thanks & regards,
Jagadish.

Dear Jagadish.

I selected the pin corresponding to SPI4 as shown below in Halcogen's PINMUX tab.

(K18) Select SPI4_CLK
(U1) Select SPI4_NCS0
(V6) Select SPI4_SOMI
(W5) Select SPI4_SIMO

And I pulled the wire from the pin corresponding to SPI4 from the J11 connector.
HET100, HET103, HET105, HET102

Can I connect like this?

Could you please let me know if there is anything more I need to do?

Best Regards,

IBLEE

Hi IBLEE,

Yes, you are correct.

(K18) Select SPI4_CLK       -   HET100
(U1) Select SPI4_NCS0      -   HET103
(V6) Select SPI4_SOMI       -   HET105
(W5) Select SPI4_SIMO      -    HET102

And make sure to OFF the USB_HOST1_ON dip switch.

--
Thanks & Regards,
Jagadish.

Dear Jagadish.

"make sure to OFF the USB_HOST1_ON dip switch."

Where do I turn the USB_HOST1_ON dip switch OFF?

Best Regards,

IBLEE

Dear Jagadish.

Is the USB_HOST1_ON dip switch talking about S2 on the devkit?

Then S2 is turned off by default.

Best Regards,

IBLEE

Hi IBLEE,

Second switch(USBH1 ON) in S2 DIP switch. Make sure to OFF this.

--
Thanks & regards,
Jagadish.

Dear Jagadish.

Thank you so much for your kind explanation.

I tested SPI4 with the settings as above.

I set devkit1 as SPI4 Master, set devkit2 as SPI4 Slave, and connected the SPI4 wire.

The test results show that the slave receives the data sent from the master well, but the data sent from the slave is only received as 0x00 by the master.

Even if you take a picture of the MOSI pin with an oscilloscope, you cannot see the data sent by the slave...

How should I debug this?
Can you help me?

Best Regards,

IBLEE

Hi IBLEE,

INBUM LEE said:

Even if you take a picture of the MOSI pin with an oscilloscope, you cannot see the data sent by the slave...

Isn't SOMI? you said slave sending data not receiving properly on master right?

Can you please send your master and slave projects for the verification?

Dear Jagadish.

oh! I said it wrong. The pin I took with the oscilloscope is the MISO (HET105) pin.

I can't send you Master/Slave Project sources.
The project I worked on is a ported source of safeRTOS, and the one I worked on is a source that created a task and registered the GIO pin as an IRQ to enable SPI Read/Write when an IRQ occurs. My boss says this cannot be shared for security reasons.

If you can't share the source, you can't help anymore, right?

Best Regards,
IBLEE

Hi IBLEE,

Can you please verify below SPI slave example once and verify the HALCoGen configurations according to this once.

(+) [FAQ] TMS570LC4357: Examples and Demos available for Hercules Controllers (E.g. TMS570x, RM57x and RM46x etc) - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums

--
Thanks & regards,
Jagadish.

Dear Jagadish.

There was a difference.
All of the SPI4 Slave Pins I set were set to Pull down.
However, the example source you gave is all set to Pull Up.

Even if I set everything to Pull Up and test it according to the example source you gave, the result is the same.

What should I do now?

Best Regards,
IBLEE

Dear Jagadish.

I did some more testing.

Currently, when sending data from Master to Slave, it is received normally, so I took a picture of the waveform with an oscilloscope when sending data from Master to Slave.

But I see something a little strange. Could you please check if it's correct?

Clk, CS0, and SIMO show normal waveforms, but the SOMI (HET105) waveform moves exactly like CS0.

In the attached photo, the yellow color is SOMI (HET105), and the blue color is CS0 (HET103).

Is it right to move like this?

Best Regards,

IBLEE

Dear Jagadish,

When I tested further, I found that the data received from the Master was not 0x00, but was received brokenly.

When the baud rate is set to 3M, the Slave receives it normally, but the Master receives it with some corruption.

If I set the baudrate to 2M and run the same test, the Slave is still receiving normally, and the Master is receiving well, but after a certain period of time, the data is received corrupted.

How should I debug this?

Best Regards,

IBLEE

Hi IBLEE,

(+) RM46L852: Arm-based microcontrollers - INTERNAL forum - Arm-based microcontrollers - INTERNAL forum - Arm-based microcontrollers - INTERNAL - TI E2E support forums

Very recently, in above thread i just created SPI-Master and Slave codes and tested without any problem. If possible, can you please also refer it.

--
Thanks & regards,
Jagadish.

Dear Jagadish.

The link page above says not found.
Could you tell me again?

Best Regards,

IBLEE

Hi IBLEE,

I converted thread into PDF:

Thread.pdf

Here are the projects i attached in above thread:

6675.SPI_MASTER_TEST_LC4357.zip

SPI_Slave_TEST_RM46.zip

--
Thanks & Regards,
Jagadish.

Dear Jagadish.

The difference between the source you shared at Halcogen and mine is that clock Polarity is Enabled and the Pins used in the SPI Slave are Pull-Up.
And in the source code, I can't find any difference other than that the CS_HOLD value is FLASE.

Currently, there is no problem in communicating with SPI3 and 5 even if the baud rate is raised to 3M, but if the baudrate for SPI4 is raised to 3M, the received value is randomly broken in the middle. (When using 1M, there was no problem for 3 days)

The Het driver is also enabled, so I am debugging, imagining whether there is interference between the Het driver and the SOMI Pin of SPI4.

Share my current situation.

Best Regards,

IBLEE

Hi IBLEE,

INBUM LEE said:

The difference between the source you shared at Halcogen and mine is that clock Polarity is Enabled and the Pins used in the SPI Slave are Pull-Up.
And in the source code, I can't find any difference other than that the CS_HOLD value is FLASE.

I enabled clock polarity because the corresponding customer said he is getting issue when clock polarity enabled. So, i enabled clock polarity on the both the ends master and slave and tested it. But in your case, it may not require.

INBUM LEE said:

Currently, there is no problem in communicating with SPI3 and 5 even if the baud rate is raised to 3M, but if the baudrate for SPI4 is raised to 3M, the received value is randomly broken in the middle. (When using 1M, there was no problem for 3 days)

My suggestion would be why don't you create one new simplest project with SPI4, and make sure to not use any other peripherals and this will give an idea how SPI4 working. SPI should work for 25Mhz without any issues.

--
Thanks & Regards,
Jagadish.

Dear Jagadish.

I'll do whatever you ask.
But I think it will take some time. I'm busy with what I'm currently doing.

If I have any questions, I'll ask again.
thank you.

Best regards,
IBLEE