fix sanity checks for new CFG_sx1276_radio boards#790
fix sanity checks for new CFG_sx1276_radio boards#790croghostrider wants to merge 1 commit intomcci-catena:masterfrom
Conversation
|
Since they changed the silicon ID, I conclude that there are other changes we have to take into account. The LMIC should record the silicon ID for easy reference at this point; and we must review the LMIC to determine whether other changes are needed. See, for example, #787, which is possibly related. I'd rather not make this change until the review has been done; and I have no time to review, nor hardware that has this value in the registers. Deferring. |
Turns out, not related. The datasheet with the change still shows 0x12 as the chip id. Google search indicates that the differences between 0x12 and 0x13 are not yet documented. More info is needed. |
|
I checked today on Semtech's site and there's still not an updated datasheet with the new chip ID. |
|
A detailed review and comparison of the latest datasheet shows a new footnote: |
| u1_t v = readReg(RegVersion); | ||
| #ifdef CFG_sx1276_radio | ||
| if(v != 0x12 ) | ||
| if(v != 0x12 & v != 0x13) |
There was a problem hiding this comment.
if(v != 0x12 & v != 0x13):
This should be &&, not &. Or else if ((v & ~1u) == 0x12), which might be smaller on AVR. Ideally, this would be a macro or inline check: if (is_sx1276_like(v)), where is_sx1276_like() looks at v and makes the decision.
We've not heard more on this since the last few years; has Semtech documented this? Or is this not relevant after all?
2 participants
No description provided.