Question from the Customer:I am using the Aardvark SPI/I2C Host Adapter as an SPI slave simulator, using a python script to simulate the behavior of a slave sensor that is connected to our main MCU board. The script I am using is aaspi_slave.py, which is included in the Aardvark Software API package. Here is a summary of my observations and the results:
The simulation program manages through the Aardvark adapter to receive the SPI master requests successfully. For each request, the Aardvark adapter must send an answer, for which I use the function:
sentCount = aa_spi_slave_set_response(handle, LTCresponse)
However, the master never receives the information that the Aardvark slave simulator sends. When I probe the MISO line, I see the line transmits some nonsensical data, which is not what was sent from the Aardvark slave simulator.
I manage to receive and display packets sent from the master device correctly. However, sending the response from the Aardvark adapter to the master device does not work. I get incorrect data packets on the MISO link:
LTCresponse = (array('B',[255,255,255,124,106,124,106,124,106, pec0, pec1]),11)
sentCount = aa_spi_slave_set_response(handle, LTCresponse)
My observation and interpretation:
- The call for that set response function sends the array of data within the MISO line at the next master cycle, but that data does not appear on the MISO line.
- Instead, a combination of a duplicate MOSI data and some other data is displayed on the MISO line. Note: there is no electrical contact between MISO and MOSI.
Are there any special considerations or settings to make sure the Aardvark configured as an SPI slave can send responses properly?
Response from Technical Support:Thank you for your questions! As is often the case, details are needed to troubleshoot and resolve a technical issue. To help address your problem, we are providing information about applying slave simulation with Aardvark adapter and how to further analyze the setup.
How an SPI Slave Simulator Works with Data
For correct transmission from the slave device to the master device, there must be at least 4 µs delay between the end of byte n and start of byte n+1. This enables setting up MISO data correctly for byte n+1. Otherwise, the Aardvark adapter will simply return the data it just received.
For the slave device to receive data, at least 10 µs are required between the start of byte n and the start of byte n+1.
When the Aardvark adapter is configured as an SPI slave, and the slave select is pulled high to indicate the end of a transaction, there is a data processing overhead of sending the transaction to the PC host. In this case, if the SPI master sends a subsequent transaction in rapid succession to the Aardvark slave, the data received by the Aardvark slave may be corrupted. There is no precise value for this minimum inter-transaction time; our suggestion is 100-200 µs.
Troubleshooting Tips
We have two recommendations: verify the signal characteristics and verify the setup of the Aardvark adapter as an SPI slave.
Signal Characteristics
It is possible that your target system, the SPI master device, may not follow the Aardvark SPI slave characteristics requirements. In this scenario, trace the signals, and if necessary, modify the target system (SPI master), to fulfill the Aardvark slave requirements. For more information, refer to Section 2.4 SPI Signaling Characteristics in the Aardvark user manual.
Loop Test with a Second Host Adapter
In addition to the aaspi_slave API script, use aaspi_file with a second Aardvark adapter. This way, you can verify the SPI slave functionality of the Aardvark device:
- Run aaspi_slave with the first Aardvark device to wait for a new slave transmission.
- In another shell, run aaspi_file to transmit a binary file with the second Aardvark device.
We hope this answers your questions. Additional resources that you may find helpful include the following:
If you want more information, feel free to contact us with your questions, or request a demo that applies to your application.