ST25RU3993 emits a frequency that differs from the requested one by 100kHz. Why could that happen?
We built a system based on application notes and evaluation board schematics and design. It consists of the MCU (stm32f407), RFID reader ST25RU3993, and some RF-related stuff. ST25RU3993 is clocked by an external 20MHz crystal oscillator.
Datasheet for ST25RU3993 says that you can select frequency by selecting reference frequency (25, 50, 100, or 125 kHz) and multipliers *33 and *32 - register A and register B respectively.
So, the final formula for the Frequency is F = Fref * (REG_A*33 + REG_B*32).
Also, they packed in 3 registers, namely PLL MAIN REGISTER 1, 2, 3 (addresses are 17h, 18h, 19h).
We observe the strange behavior of our system, mentioned in the main question of this message. I used a "signal hunter" device, which is actually an air spectrum analyzer, in order to evaluate, which is actual frequency emitted by the system. And it differs from the requested one. Namely, we requested 865.7 MHz, but the frequency analyzer detects a signal on 865.8 Mhz (see picture: center is on the 865.7MHz, current peek frequency written in yellow at the top of the screen 865.8MHz).
In these cases the following register values were used:
For 25kHz Fref A=548 B=517 regs: 78h 16h 24h
For 50kHz Fref A=258 B=275 regs: 64h 4Dh 02h
For 100kHz Fref A=145 B=121 regs: 51h E4h 91h
I have double-checked the values, written to the registers, and even read them back. Their values look correct. So it looks like the system has some kind of erroneous behavior source in its design.
This behavior does not depend on the reference frequency and has been reproduced with 25,50 and 100 kHz reference frequencies. It also happens at different frequencies: 865.7 MHz, 866.3MHz, 866.9MHz, 867.5MHz.
Any support from ST engineers is greatly appreciated. Can you reproduce such behavior on your side?
Best regards.
Alexander.
