Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add a function to activate CW mode #16

Merged
merged 5 commits into from
May 3, 2023
Merged

Conversation

Miceuz
Copy link
Contributor

@Miceuz Miceuz commented Mar 15, 2023

It's just a wrapper function for underlying Semtech infrastructure for activating CW mode.

@Miceuz Miceuz mentioned this pull request Mar 15, 2023
Copy link
Collaborator

@matthijskooijman matthijskooijman left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks!

Overall, this PR looks good. Some remarks/questions:

  • You've also included a commit for RAK support, that probably warrants its own PR. Could you remove it (just force push to your branch with the commit removed).
  • Are you sure the power is directly specified in dB? The regular TX power is specified by index, with a non-trivial conversion from dB:
    bool STM32LoRaWAN::powerdB(int8_t db){
    // This uses knowledge about the radio implementation to calculate the
    // index to use. See RegionCommonComputeTxPower()
    int8_t index = -(db - 1) / 2;
    return mibSetInt8("power", MIB_CHANNELS_TX_POWER, index);
    }
  • Is there a special "no timeout" value? If so, the comments should specify it.

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 15, 2023

Argh, you are right, I have based my changes on previous branch. Will fix that.

You are right about power units also.

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 15, 2023

Fixed that

@Miceuz Miceuz requested a review from matthijskooijman March 15, 2023 12:02
@matthijskooijman
Copy link
Collaborator

Seems you dropped the RAK commit, but haven't addressed the other questions about power and timeout yet?

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 15, 2023

Seems you dropped the RAK commit, but haven't addressed the other questions about power and timeout yet?

Yes, I have removed dBm mention from the documentation and no, there is no special "no timeout" value it seems:

UTIL_TIMER_Status_t UTIL_TIMER_SetPeriod(UTIL_TIMER_Object_t *TimerObject, uint32_t NewPeriodValue)
{
  UTIL_TIMER_Status_t  ret = UTIL_TIMER_OK;
  
  if(NULL == TimerObject)
  {
	  ret = UTIL_TIMER_INVALID_PARAM;
  }
  else
  {
    TimerObject->ReloadValue = UTIL_TimerDriver.ms2Tick(NewPeriodValue);
    if(TimerExists(TimerObject))
    {
      (void)UTIL_TIMER_Stop(TimerObject);
      ret = UTIL_TIMER_Start(TimerObject);
    }
  }
  return ret;
}

Time is converted from seconds to ms in radio.c:

    uint32_t timeout = ( uint32_t )time * 1000;
    uint8_t antswitchpow;

    SUBGRF_SetRfFrequency( freq );

    antswitchpow = SUBGRF_SetRfTxPower( power );

    /* WORKAROUND - Trimming the output voltage power_ldo to 3.3V */
    SUBGRF_WriteRegister(REG_DRV_CTRL, 0x7 << 1);

    /* Set RF switch */
    SUBGRF_SetSwitch( antswitchpow, RFSWITCH_TX );

    SUBGRF_SetTxContinuousWave( );

    TimerSetValue( &TxTimeoutTimer, timeout );
    TimerStart( &TxTimeoutTimer );

@fpistm fpistm added the enhancement New feature or request label Mar 23, 2023
Copy link
Member

@fpistm fpistm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Small typos else LGTM.
Thanks for your contribution.

src/STM32LoRaWAN.h Outdated Show resolved Hide resolved
src/STM32LoRaWAN.h Outdated Show resolved Hide resolved
@fpistm fpistm added this to the 0.1.1/0.2.0 milestone Mar 23, 2023
@matthijskooijman
Copy link
Collaborator

Yes, I have removed dBm mention from the documentation and no, there is no special "no timeout" value it seems:

I wonder if the API should accept a power in dBm, just like the regular power setting function. Of course it is a testing function, so having a somewhat less friendly API that has a potentially hardware-dependent power argument might be acceptable (@fpistm, what do you think?), but in that case at least the documentation should say something about what the power values mean (even if it is just a reference to the relevant register in the STM32WL55 datasheet).

@fpistm
Copy link
Member

fpistm commented Mar 23, 2023

Yes, I have removed dBm mention from the documentation and no, there is no special "no timeout" value it seems:

I wonder if the API should accept a power in dBm, just like the regular power setting function. Of course it is a testing function, so having a somewhat less friendly API that has a potentially hardware-dependent power argument might be acceptable (@fpistm, what do you think?), but in that case at least the documentation should say something about what the power values mean (even if it is just a reference to the relevant register in the STM32WL55 datasheet).

Agreed.
* \param power TX power level [0-15]. is confusing as it seems it should be in dBm.
So I would rewrite this: * \param powerdBm transmit power in dBm. and rename arg powerdBm

Miceuz and others added 2 commits March 23, 2023 19:05
Fix a typo

Co-authored-by: Frederic Pillon <frederic.pillon@st.com>
Signed-off-by: Albertas Mickėnas <albertas@technarium.lt>
Fix a typo

Co-authored-by: Frederic Pillon <frederic.pillon@st.com>
Signed-off-by: Albertas Mickėnas <albertas@technarium.lt>
@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 23, 2023

Yes, I have removed dBm mention from the documentation and no, there is no special "no timeout" value it seems:

I wonder if the API should accept a power in dBm, just like the regular power setting function. Of course it is a testing function, so having a somewhat less friendly API that has a potentially hardware-dependent power argument might be acceptable (@fpistm, what do you think?), but in that case at least the documentation should say something about what the power values mean (even if it is just a reference to the relevant register in the STM32WL55 datasheet).

Agreed. * \param power TX power level [0-15]. is confusing as it seems it should be in dBm. So I would rewrite this: * \param powerdBm transmit power in dBm. and rename arg powerdBm

Cool, I will re-calculate to index from dB in the same manner you have posted above int8_t index = -(db - 1) / 2;

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 23, 2023

I found something fishy in Semtech code. SUBGRF_SetRfTxPower declares as if it expects an index, not dBm, that's what initially threw me off:

uint8_t SUBGRF_SetRfTxPower(int8_t power)
Set the Tx End Device conducted power

Parameters:
[in] – power Tx power level [0..15]

Return values:
paSelect [RFO_LP, RFO_HP]

Then it does checks for PA selection:

    switch (TxConfig)
    {
        case RBI_CONF_RFO_LP_HP:
        {
            if (power > 15)
            {
                paSelect = RFO_HP;
            }
            else
            {
                paSelect = RFO_LP;
            }
            break;
        }

And then calls SUBGRF_SetTxParams which expects dBm:

void SUBGRF_SetTxParams(uint8_t paSelect, int8_t power, RadioRampTimes_t rampTime)
Sets the transmission parameters

Parameters:
[in] – paSelect RegPaConfig PaSelect value (RFO_LP, RFO_HP, etc)
[in] – power RF output power [-18..13] dBm
[in] – rampTime Transmission ramp up time

Thus the check for 15 in SUBGRF_SetRfTxPower makes no sense and also [in] – power Tx power level [0..15] is wrong as SUBGRF_SetTxParams expects [-18..13] dBm.

SUBGRF_SetRfTxPower is called also in RadioSetTxConfig that declares it expects power in dBm, and that is called by RegionEU868TxConfig where power index is actually converted to dBm by RegionCommonComputeTxPower.

So yeah, I think SUBGRF_SetRfTxPower needs to be fixed.

Can anyone guide me on >15 check?

@fpistm
Copy link
Member

fpistm commented Mar 24, 2023

Looking at the code, when LoRaMacMlmeRequest is called, it calls:

status = SetTxContinuousWave1( mlmeRequest->Req.TxCw.Timeout, mlmeRequest->Req.TxCw.Frequency, mlmeRequest->Req.TxCw.Power );

then SetTxContinuousWave:
Radio.SetTxContinuousWave( frequency, power, timeout );

with SetTxContinuousWave :
/*!
* \brief Sets the radio in continuous wave transmission mode
*
* \param [in] freq Channel RF frequency
* \param [in] power Sets the output power [dBm]
* \param [in] time Transmission mode timeout [s]
*/
void ( *SetTxContinuousWave )( uint32_t freq, int8_t power, uint16_t time );

So I think we could only deal with dBm while it is explicitly describe, as this feature is for test purpose, I think it is acceptable.

@matthijskooijman ?

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 24, 2023

Yes, and then RadioSetTxContinuousWave calls SUBGRF_SetRfTxPower that expects Tx power level [0..15] that calls SUBGRF_SetTxParams that expects RF output power [-18..13] dBm.

So it's clearly dBm, but I think there's a bug in SUBGRF_SetRfTxPower - it will never select RFO_HP

@fpistm
Copy link
Member

fpistm commented Mar 24, 2023

Yes, and then RadioSetTxContinuousWave calls SUBGRF_SetRfTxPower that expects Tx power level [0..15] that calls SUBGRF_SetTxParams that expects RF output power [-18..13] dBm.

So it's clearly dBm, but I think there's a bug in SUBGRF_SetRfTxPower - it will never select RFO_HP

Right, I guess an issue should be filled in https://github.com/STMicroelectronics/STM32CubeWL/issues

@Miceuz
Copy link
Contributor Author

Miceuz commented Mar 24, 2023

Yes, and then RadioSetTxContinuousWave calls SUBGRF_SetRfTxPower that expects Tx power level [0..15] that calls SUBGRF_SetTxParams that expects RF output power [-18..13] dBm.
So it's clearly dBm, but I think there's a bug in SUBGRF_SetRfTxPower - it will never select RFO_HP

Right, I guess an issue should be filled in https://github.com/STMicroelectronics/STM32CubeWL/issues

Ok, will do. So this PR then can be merged - I have altered the documentation strings to represent dBm.

Copy link
Member

@fpistm fpistm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Thanks for the update.
Waiting @matthijskooijman feedback 😉

@fpistm
Copy link
Member

fpistm commented Apr 11, 2023

Thanks @Miceuz to the issue submitted here: STMicroelectronics/STM32CubeWL#64

@matthijskooijman are you agreed with this PR now? I guess it could be merged safely as it does not impact other API.

@fpistm fpistm modified the milestones: 0.1.1, 0.1.2/0.2.0 Apr 11, 2023
Copy link
Collaborator

@matthijskooijman matthijskooijman left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I was a bit confused by all the different code snippets, but finally found the time to look a bit closer. Looks good to me now, yes.

As for SUBGRF_SetRfTxPower(), AFAICS the problem is not actually the code, just the documentation problem, right? Reading your comments and the code right, the function is called with dBm, passes dBm on to SUBGRF_SetTxParams() and the >15 check makes sense, since LP goes up to 15dBm and HP up to 22dBm, so it should work as expected, right?

For reference, here is the upstream issue @Miceuz reported: STMicroelectronics/STM32CubeWL#64

@fpistm
Copy link
Member

fpistm commented May 3, 2023

Agreed, documentations are not coherent. Let's see how it will be handled on STM32CubeWL repo.
Anyway, this PR can be merged safely.

@fpistm fpistm merged commit d7dbe30 into stm32duino:main May 3, 2023
@fpistm fpistm linked an issue May 3, 2023 that may be closed by this pull request
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

Successfully merging this pull request may close these issues.

CW mode support?
3 participants