Repository navigation
TX power calculation assumes 100% power is 100w #7
Description
I applied the patch offered here to properly calculate power, but the result was that my QRP IC-705 is said to produce 50w, not 5!
My quick hack was to replace data.Power = powerPercent * 100 to replace data.Power = powerPercent * 10 Perhaps there is a more global solution.
Activity
johnsonm commentedon Aug 29, 2026
Please try the v0.1.0 release and see if it is correct for you?
If not, we might need config for a power scaling factor. 🤔
johnsonm commentedon Aug 29, 2026
Or later. I just released 0.2.0
I'm now subscribed to updates in my own project (thanks, GitHub, for assuming that I didn't want to get updates about my own project!) so I'll actually see if you have any feedback now. 😳
johnsonm commentedon Aug 30, 2026
If v0.2.0 didn't fix this, then I hope that v0.3.0 with more extensive hamlib PWR fixes from @int2001 does. Please download and let me know!
brobertson commentedon Sep 7, 2026
Using both the supplied binary and a git clone, I get errors that seem to relate to pwr not being reported as an integer. Moreover, the 100% seems still to be assumed to be 100w:
Sep 07 13:29:23 hampi waveloggoat[2253]: time="2026-09-07T13:29:23-03:00" level=info msg="Sending to http://ve9qr.ca/index.php/api/v2/radio: {"radio":"IC-705","power":34.117599999999996,"frequency":7029080,"mode":"CW"}"
Sep 07 13:29:23 hampi waveloggoat[2253]: time="2026-09-07T13:29:23-03:00" level=error msg="Error posting to Wavelog: wavelog API returned non-200 status code: 400. Body: {"error":{"code":"validation_error","message":"Invalid field(s): power","details":{"invalid":["power"]}}}"
brobertson commentedon Sep 7, 2026
Here's a diff file that got mine working again:
data.Power = powerPercent * 10 // Convert percentage to watts (assume 100W max) BGR: assume 10W for IC-705and
payload.Power = int(data.Power)in the else block. I'm running v3.2.0
johnsonm commentedon Sep 7, 2026
Thank you for testing! This is the downside of me currently running this only with a 100W rig. 🤣 And also still running 3.1.0, haven't upgraded yet to 3.2.0.
See PR #8 and the associated regression against 3.2.0 wavelog/wavelog#3803 — where it looks like floats are being added back for watts so that QRPp power levels can be reported accurately in 3.2.1.
Looks like I should add configuration for a power fraction for a radio to make this generic. I can imagine that if someone has an external linear amp set up they need to use a multiplier as well.
brobertson commentedon Sep 7, 2026
I agree that full integers is really not good power recording habit. I, for instance, do all my FT-8 contacts in milliwatts, usually at 0.9w. If we round to integers, I loose the distinction between 900mW and 1W, which matters a great deal to me.
Not exactly on topic, but you might want to recommend that Linux folks compile with CGO_ENABLED=1. This forces Go to use the standard C library's getaddrinfo(), which properly respects mDNS/Avahi for .local hostnames. The Go library seems not to, and lost track of the machine with rigctrld on it because I power cycled it during all this.
johnsonm commentedon Sep 7, 2026
Yeah, the wavelog bug not accepting floats is being fixed. (Even with a complaint that power levels aren't interesting...)
I intentionally don't compile the Linux binary with CGO_ENABLED=1 to avoid dependencies on the glibc version.
Can you please file another feature request for mDNS support for .local hostnames? This can be done in pure Go.
The next release, coming shortly, will enable the new
-max-powersetting that will scale the reported power.johnsonm commentedon Sep 7, 2026
v0.4.0 now released contains this fix
johnsonm commentedon Sep 7, 2026
I'll do the feature request for .local since I think I've implemented it...
johnsonm commentedon Sep 7, 2026
See #10 for mDNS