Part Number: AM3354
Other Parts Discussed in Thread: TPS61165
Tool/software:
Hi TI Experts,
Customer is using U14 pin EHRPWM1A in AM335 to drive the TPS61165 LED driver.
The original frequency they got is 4kHz shown below.
Now they need to increase this frequency to 5kHz or an even higher value.
I have proposed the following method for customer to try.
Firstly the base address of ePWM1 is 0x48302200 shown below.
Then the time-based period register in ePWM1 has offset 0xA shown below.
So I let customer read the register (0x48302200 + 0xA = 0x4830220A) firstly shown below to see the default frequency. However, customer found the result for each read is different shown below. (the first 16 bits are the same but the other bits keep different)
first read: 0x61A740A8
second read: 0x61A70EA4
This is strange to us, according to below register description, the first 16 bits stand for the PWM frequency, so we just consider the 0x61A7 and ignore the rest.
According to the calculator, the 0x61A7 is 24999, and we do not understand how it relates to the 4kHz we got using the "cat" command mentioned above.
But we have tried an experiment to change the value 0x61A7 to 0x07D0 shown below, and see if there is any visible change on the LED side.
The result shows that the LED indeed become much more brighter than before.
However, as you could see from the above picture, the result of "cat" command still shows that the frequency is still 4kHz even if LED becomes brighter.
Hence, may I have your help for the below 2 questions please?
1: How does the TBPRD register value relate to the actual frequency on the LED? (what is the mathematical relationship)
2: Why the "cat" command still show 4kHz frequency but we do see a brighter LED after changing TBPRD register value?
Thanks,
Kevin
