• AVR Freaks

Hot!RN4678 tricks and debunking

Author
aurelienr
Starting Member
  • Total Posts : 55
  • Reward points : 0
  • Joined: 2006/05/28 06:27:02
  • Location: France
  • Status: offline
2020/05/19 05:45:30 (permalink)
5 (4)

RN4678 tricks and debunking

Hello,
I've integrated the RN4678, and it took be a lot of time to make it work compared to other modules I worked with in the past.
As you can see on other topics, the documentation is full of errors and approximations. The technical support is also a nightmare, looks like they don't know the product too (but they talk like they know it). So I dressed a "small" list of informations and tricks to know the limitations. All of this is based on the latest soft relevase V1.13 at time of writing, powered with a regulated 3.8V rail.
 
- At start up after reset before device is ready, some GPIO that should be inputs (flow control for example) are set during some milliseconds as outputs with "1" level. So if you set your own GPIO as output at low level for example, it will be create a huge current draw. Wait some delay and for VIO output to be stable to turn your GPIO as outputs.
 
- The power consumption. I use the module for the moment in BT Classic mode only. The datasheet gives a consumption of 2.5mA idle not connected. But in fact, with default configuration, it consumes about 15mA ! Easy to test directly on the pictail board, just don't connect USB cable and power the connector with only 5V and GND, and compare consumption with module on and off. Modifying sniff parameters do not have impact on the consumption. When connected, the consumption is still...15mA ! I'm still dealing with the tech support to know the conditions of their measurements.

- In the command manual, P32 is defined as "unassociated". In the datasheet it is associated to LINK_DROP_CONTROL input. That can explain automatic connection abort if this IO is not correctly polarized.
 
- The exact role of SW_BTN and WAKE_UP is poorly documented. The SW_BTN is presented as "disabled defaut feature". But in fact it is mandatory to set this pin high (connect to BAT) to power on the module.
 
- If, like me, you want to be able to update fiwmare directly on your board, and if you do that through a bridge implemented into your MCU, please know that the RN4678 UART config is 115200bps 8N1 WITHOUT hardware flow control (different from app mode where flow control is enabled by default).
 
- There is a mistake in the firmware update document about EAN,P20 and P24 signals. Please follow the datasheet instead (the update guide is only valable for the pictail board).

Aurelien
#1

7 Replies Related Threads

    ric
    Super Member
    • Total Posts : 27595
    • Reward points : 0
    • Joined: 2003/11/07 12:41:26
    • Location: Australia, Melbourne
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/19 13:32:01 (permalink)
    5 (1)
    "Debunking" is strangely apt, but I think you intended "debugging". :)
     

    I also post at: PicForum
    Links to useful PIC information: http://picforum.ric323.co...opic.php?f=59&t=15
    NEW USERS: Posting images, links and code - workaround for restrictions.
    To get a useful answer, always state which PIC you are using!
    #2
    aurelienr
    Starting Member
    • Total Posts : 55
    • Reward points : 0
    • Joined: 2006/05/28 06:27:02
    • Location: France
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/19 14:55:35 (permalink)
    5 (1)
    When you spend several days on that kind of problems, no word is strong enough :)
    #3
    Francesco83
    New Member
    • Total Posts : 3
    • Reward points : 0
    • Joined: 2020/05/26 12:52:41
    • Location: 0
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/27 07:34:05 (permalink)
    0
    I'd like to implement this module in a project so I've bought the RN4678 PICtail board  to evelaute it.
    Here's the documentation I've used:
     
    (RN4678 PICtail board)
    https://www.microchip.com/DevelopmentTools/ProductDetails/PartNO/RN-4678-PICTAIL
     
    (RN4678 Datasheet)
    http://ww1.microchip.com/downloads/en/DeviceDoc/50002519E.pdf
     
    (Command reference user guide)
    http://ww1.microchip.com/downloads/en/DeviceDoc/RN4678-Bluetooth-Dual-Mode-Module-Command-Reference-User-Guide-DS50002506C.pdf
     
    Since I was not satisifed with the speed at 115.200 baud rate (I'm trying to use it to send datas from a sensor at 600Hz through a MCU connected to the RN4678 via UART and at 115.200 it loses characters) and it is stated that the best transfer speed using this module is obtained setting the UART at 1Mbit I've tried to use the command:
    SU,11 
    to change the baud rate to 1Mbit.
    At first it wasn't accepeted but I've found out it was because of an old version of the firmware preloaded on the board. So I've updated to v1.13.
    Now the command SU,11 is accepted but even after a reboot it seems like the UART simply doesn't work.
    If I use TeraTerm with the baud rate set a 1Mbit and I press reset on the evaluation board usually the string %Reboot% is sent over the UART but at 1Mbit nothing appears at all. Usually if there's a mismatch in the baudrate characters still appear but badly interpreted, in this case nothing appears, it's as if the UART simply doesn't send datas.
    Could you help me with this? Is the PICtail board tested for 1Mbit baud rates or it's just the module that supports that baud rate?
     
    I also must say that the documentation for this module has many contradictions and missing things.
    It is stated that hardware flow control is enabled by default but it turns out that the PIC tail board works perfectly out of the box with hardware control disabled in TeraTerm, it is also stated that hardware flow control can be disabled but I've not found any command in the reference guide posted above to actually disable it!
    On page 16 there's a command described as SQ,1000 to enable it but it's not stated how to disable it.
    I guess it can be forced by putting CTS to GND anyway? But the documentation is not clear.
     
    Also last point is the operating mode setting using the three pins P2.0, P2.4 and EAN.
    In the datasheet it is stated that EAN must be pulled down to GND through a 4.7ohm resistor on page 4.
    Then on page 6 it is stated that all three pins are internally pulled up.
    Then in the reference circuit on page 12 all the three pins are directly connected to GND through jumpers.
    Finally in the PICtail board the three pins are controlled in a totally different way! EAN is actually pulled up to 3.3V if a dip switch is closed (why a pullup is needed if the pin has an internal pullup??) while P2.0 and P2.4 are pulled down to GND through 4.7Kohm resistors (that were not present in the reference circuit in the datasheet).
    Honestly what's up with all these differences? I guess that since the PICtail board actually works the truth is in the reference PCB but this confusion is frustrating if you want to design something around this product.
     
    Could you help me with these issues? Particularly the one with the 1Mbit baud rate not working?
    Where can I ask for more specific technical support in case?
    Thank you!
     
    #4
    Francesco83
    New Member
    • Total Posts : 3
    • Reward points : 0
    • Joined: 2020/05/26 12:52:41
    • Location: 0
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/27 07:35:16 (permalink)
    5 (1)
    I've posted here because it seems I can't create a new dedicated thread... I get an access denied error???
    #5
    aurelienr
    Starting Member
    • Total Posts : 55
    • Reward points : 0
    • Joined: 2006/05/28 06:27:02
    • Location: France
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/28 05:21:18 (permalink)
    0
    Francesco,
    I'm sorry the 115200bps CTS/RTS config is ok for me I did not try to increase comm speed or flow control.
     
    About EAN, in my case I don't have pull up or pulldown, this is directly driven by my MCU. The datasheet indicates 4.7KR pull down, not 4.7R. On the pictail this is different : it is connected to VCC through a 10R resistor (why ?), or let floating (pull down including in the module) ?
    It looks like the module datasheet is a simple copy of the IS1678 chip inside (also produced by microchip)....so it's possible that the pull down is integrated into the module, so it appears in the IS1678 doc but has been copied to the module doc...
     
    I guess we will have the final answer within some years, time it tooks to microchip to upgrade the doc of their RTC family MCP7951x full of bugs. At the beginning, the tech support was responding bullshit, until an erratum has been published, confirming my observations...(personnel advice : go away from this family :D)
     
    Good luck !
     
    Aurélien
     
     
    #6
    Francesco83
    New Member
    • Total Posts : 3
    • Reward points : 0
    • Joined: 2020/05/26 12:52:41
    • Location: 0
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/31 05:02:13 (permalink)
    0
    aurelienr
    Francesco,
    I'm sorry the 115200bps CTS/RTS config is ok for me I did not try to increase comm speed or flow control.
     
    About EAN, in my case I don't have pull up or pulldown, this is directly driven by my MCU. The datasheet indicates 4.7KR pull down, not 4.7R. On the pictail this is different : it is connected to VCC through a 10R resistor (why ?), or let floating (pull down including in the module) ?
    It looks like the module datasheet is a simple copy of the IS1678 chip inside (also produced by microchip)....so it's possible that the pull down is integrated into the module, so it appears in the IS1678 doc but has been copied to the module doc...
     
    I guess we will have the final answer within some years, time it tooks to microchip to upgrade the doc of their RTC family MCP7951x full of bugs. At the beginning, the tech support was responding bullshit, until an erratum has been published, confirming my observations...(personnel advice : go away from this family :D)
     
    Good luck !
     
    Aurélien
     
     




    Thanks for the reply Aurélien, unfortunately I still haven't been able to understand what's the problem with the 1Mbit setting not working on my module since the relative command seems to be accepted by the board.
    As you can see here on the message boards there's very little support and active people among Microchip's employees so it's disappointing.
    I guess that I have to ask for a ticket to get support?
     
    #7
    ric
    Super Member
    • Total Posts : 27595
    • Reward points : 0
    • Joined: 2003/11/07 12:41:26
    • Location: Australia, Melbourne
    • Status: offline
    Re: RN4678 tricks and debunking 2020/05/31 05:12:54 (permalink)
    0
    Francesco83
    ...
    As you can see here on the message boards there's very little support and active people among Microchip's employees so it's disappointing.
    I guess that I have to ask for a ticket to get support?

    That is how you ask for official support from Microchip.
    This board is just a user forum.
     

    I also post at: PicForum
    Links to useful PIC information: http://picforum.ric323.co...opic.php?f=59&t=15
    NEW USERS: Posting images, links and code - workaround for restrictions.
    To get a useful answer, always state which PIC you are using!
    #8
    Jump to:
    © 2020 APG vNext Commercial Version 4.5