Hi,
I'm using a PAN1740 HW02 / SW 02 module and have scanning and connection problems after burning the OTP. I use the newest SDK5.0.3 and the prox_monitor_ext_580.hex µVision5 project. When I use this file with the USB dongle and selecting #define HW_CONFIG (4) in the user_periph_setup.h file I can boot with connection manager as central or peripheral. I replaced the PAN1740 at the USB dongle from a HW01/ SW02 to a HW02 /SW02 because I have huge problems to connect with an iOS App to the HW02/ SW02 module. I burned the OTP and the module is advertising but when I try to connect the module I need many tries or the connection can't be established.
But when I use the connection manager to program the bin file to the dongle I can connect to the PAN1740 immediately.
With the connection manager the firmware is flashed in the SysRam?
When I use one USB dongle as master and one as Slave the connection between both can also be established immediately.
What is the difference to programming the same firmware using OTP?
Why is the connection so bad and with the old HW01 version this works.
Look at this link
http://support.dialog-semiconductor.com/pan1740-connection-problems-usin...
I have posted my problems but I get no answer. Now, I think it's a difference between SysRam and OTP because the iOS App can also connect to the USB dongle immediately.
Perhaps my OTP header is wrong, but with HW01 there are no problems with connection but there I burned the SDK3.0.10 prox_monitor_ext_usb.bin file to OTP.
Please can you help me because I have no more ideas to solve the problems.
问候
Eckart
HI,
The changes of HW_CONFIG from 0 to 4 declares no RTS and CTS signals been used in the application. Please verify your external host software been configured to use the RTX and CTS signals for flow control handshaking before starting the command exchange, or not.
如果连接成功之间的间歇and failure, please use SmartSnippets tool and JLINK to check the OTP Header content at address 0x47F8C and 0x47F78. The XTAL16MHz @ 0x47F8C should not be zero and Calibration Flag @ 0x47F78 bit 4 should be zero. If not, the crystal has not been calibration that may cause connection issue.
Moreover, please check whether 32KHz crystal is present in your module. If not, please configure CFG_LP_CLK in da1458x_config_advanced.h to LP_CLK_RCX20 to avoid the firmware go into sleep mode assumed 32KHz crystal timing.
Hope this info can help to solve your issue, If not, state briefly what you observe in the mentioed changes.
问候, WC Ip
Hi,
I used the prox_monitor_ext_580.bin file for OTP image and set the Calibration Flag @ 0x47F78 bit 4 to zero with the value of A5A5002D using SmartSnippets and burned the OTP header. But when I read the header from memory there is again the value of A5A5003D. After trying to connect the module with the iOS App I have also connection problems with the PAN1740. Nothing have changed.
I changed the attached header.hex file with the value of A5A5002D and tried to import this file in SmartSnippets. After this I get the following error:
(错误@16-01-04 15:03:39)HexFile Data Record line checksum error (line 32): :047F7800A5A5002D7E
I also get some warnings like
[WARNING @16-01-04 14:37:31] Expected trim value of RC16MHz is 0000000A while user-entered trim value is 00000008.
when burning OTP header. I ignore this and with PAN1740 HW01/ SW02 this also works.
The XTAL16MHz @ 0x47F8C is not zero and when I'm looking in the datasheet of the PAN1740 there's a 32kHz clock in block diagram so I think the module have the XTAL32kHz.
When I understand you right there's a problem with sleep mode in OTP mode? So if I try to connect the module it goes to deep sleep or someting else and there can not be a connection established? Why is that problem not in HW 01? Because every module with HW01/ SW02 works fine with the header I attached and the full_emb.bin file as OTP image. What is the difference between HW01 and HW02?
Please look about my heder.hex file if there´s any error I didn't see.
Hope you can help me.
问候
Eckart
Hi,
I have a type mistake in the following statement.
"The XTAL16MHz @ 0x47F8C should not be zero and Calibration Flag @ 0x47F78 bit 4 should "NOT" be zero."
For your case, I really don't know the difference between PAN1740 HW01 and HW02 as its data sheet doesn't describ at all.
To connect to iPHONE, you cannot use "prox_monitor_ext_580.bin" as it is central firmware. You should use "prox_reporter_xxx" instead.
The header has been inspected and the content indicated 32KHz crystal installed and the XTAL16M has been tuned though its value is quite high.
Please try the "prox_reporter" firmware and diescribe the result based on LightBlue on iPhone.
问候,
Hi,
I have tried to use the prox_reporter.hex project and burned the OTP. Afterwards I can't communicate with te PAN1740 and nothing works.
But I have found something depending on the RC16Mhz Trim Values and using full_emb.bin file. If the value is 0x000000A I can connect the module and if the value is 0x000000B I have a lot of connection problems.
So I entered a 0x000000A to this RC16Mhz Trim Values and burned the OTP header but after read the header back the value was 0x000000B and I have connection problems too.
I don't understand why this value changed while I entered 0x000000A for burning. How can I set this value and it can't change?
Can you explain me this phenomenon and if there's relationship between this Trim Value and my connection problems?
Thank you.
问候
Eckart
Hi ehanl,
The RC16MHz Trim value and VCO are calculated and burned during production so you can't change them. As far as i know some samples may suffer from wrong RC16_Trim programming and wrong VCO_Trim values. Depending on the written RC16 value, the device might work or not work. Those values are updated but you might have an old silicon chip and that might cause the problem.
Thanks MT_dialog
Hi MT_dialog,
what do you mean with old silicon chip?
问候Eckart
Hi ehanl,
There might be a chance that the chip you have is a bit old.
Thanks MT_dialog