CC2500 Tx Power test - Skyartec & Frsky Protocols

I have attempted to fix the following issue:

Changing The tx power has no effect on Range when using the Skyartec protocol. Range is always very long even at 100uw.

NOTE: This would also apply to Frsky protocols

As I am unable to test this myself, I have created a test version. If you're able to assist, please try the test version and check range at various low power levels and report back.

Test Builds

The change to patable codes is not important, I just selected lower current draw equivalents.

The patable must be loaded onto CC2500, then table is then used to ramp up the power level when powered on and ramped down when powered off. The power level is changed by 3 bits within register FREND0.

SkyArtec.patch

I agree that u8 is over used and u32 is better and often reduces firmware size...

In this particular case u8 or u32 makes no difference to firmware size because the input parameter Model.TxPower is u8. From memory there are 5 of these ?_SetPower() functions, the other 4 all use u8 so I chose u8.
- u8 involves no data type conversion.
- u32 involves a data type conversion that will have no cost on little endian.

I changed it because int is signed. Changing int to u8 removes an unnecessary conversion between signed and unsigned types.

uint_least8_t would be better but I've found that using it makes code less readable.
uint_fast8_t is good for iterator variable in loops when iterator is used in arithmetic with u8. eg. protocol code with u8 data arrays. I suggest adding to common.c
BTW: I converted dsm protocol to u16 and u32 versions but the result in both cases is an increase in firmware size so I haven't used either. I believe the increase comes from the extra bit arithmetic required to calculate the crc.

Back on topic. Did the patch work?

Or has anyone tried the test version?

yes I have. there ist no difference in range.

Thanks for testing this. I have uploaded new test builds.

This time I set power level first before enabling TX mode. So the order of these 2 commands is now reversed. The order would be important because when you enable TX mode the PA powers up (ramps up) to the pre-set level. Setting that level after powering up would have no effect.

SkyArtec_v2.patch

I hope you or someone can test this new version and report back, thanks.

Test Builds

Thanks for trying to resolve this bug!

I did a quick range check indoors, and I cant see a difference. That is the skyartec reciver dosent loose reception on 100uw long after a cyrf or a7105 reciver dose at 100uw.

Ill try to get a real range check compearing CC2500 @ 150mw to CC2500 @ 100uw in tomorrow

Thanks for testing this.
I have spotted the problem and have uploaded new Test Builds .

The lowest TX power table value was being overwritten by another piece of code.
Source code
SkyArtec_v3.patch

Oh good, my version is now working.

Could someone with Skyartec please test this and report back?

See: first post

Mine is implemented according to CC2500 datasheet .

Search that document for PA_POWER to read about how PA_TABLE is used to ramp PA_POWER up/down and decide which source you like best.


I was only trying to fix a reported bug . If the bug report is wrong then I'll just update its status.