RFM69HW high power not working for me

[JimInWoodstock]

Setup

Receiver Motino set up on PC #1 in basement. Receive sketch prints received message and RSSI level.
2 transmitter Motinos (1 RFM69W) and 1 RFM69HW) alternately  plugged into pc #2 in main level bedroom. No more 75ft distance between them.
Antennas are the ones supplied with the Motinos.
Transmitter sketch transmits a test message when enter key is hit on the keyboard.
I have the same sketch running on each transmitter with the exception of the "IS_RFM69HW" set on the RFM69HW module.
With the lower power module plugged in and the power level set to 31, I get a RSSI of around -63 at the receiver.
With the high power module plugged in and the power level set to 31, I get a RSSI of around -63 (same as above) at the receiver.
Obviously, there is something wrong. After spending much time testing this, I am at a loss.
I am attaching the send  and receive sketches. Any help would be appreciated.

[Felix]

To get an accurate measurement, everything would have to be identical, except for the variable you're measuring, which in this case is the sketch difference (one has HW set ON) and the physical difference. Even your body and other obstacles can influence this kind of reception. To get max power you should not have to call setPowerLevel().
And BTW that function was just updated.

[JimInWoodstock]

Felix,

I have the 2 modules plugged in exactly the same way. I pull the low power module off of the ftdi interface and plug in the high power module and then relaunch the serial monitor to reset the new module. So they end up in the same place physically in the bedroom. Should I reinstall the latest rfm69 library and re compile?

[Felix]

I guess exclude the setPowerLevel call completely and that should give you max output for the test.
Recompile with the exact sketch, except include the setHighPower() call for the Moteino with HW radio.

[JimInWoodstock]

I edited the transmit sketch and removed all parts that address the setpowerlevel and removed the setpowerlevel command as well.
I moved the antenna in the basement (receive system) to a better place.

I now have these RSSI readings

Low power  -55
High power -52

This is a difference, but I was expecting more than 3 points.

The new transmitter sketch is attached. I just comment out the line "radio.SetHighPower();" in the RFM69w version.

[TomWS]

I edited the transmit sketch and removed all parts that address the setpowerlevel and removed the setpowerlevel command as well.
I moved the antenna in the basement (receive system) to a better place.

I now have these RSSI readings

Low power  -55
High power -52

This is a difference, but I was expecting more than 3 points.

The new transmitter sketch is attached. I just comment out the line "radio.SetHighPower();" in the RFM69w version.

I agree that the difference seems inappropriate but, unless you disable AGC at the receiver, you're getting interactions that might not relate to your source...

Look into the LNA register values to temporarily disable AGC at your receiver.

Tom

[Felix]

Jim, really don't know what to say. Radio waves, interference, signal strength through obstacles, logarithmic signal path loss ... its much like black magic, it works but we can't really debug that. If you had a really low RSSI then maybe we had a case to study.

[JimInWoodstock]

Thanks to everyone for their comments regarding this post. So I am now moving on.

I do have another question. Does anyone know what level of signal is considered OK? I would like to know the range of RSSI for a adequate signal, good signal, excellent signal, etc. If there is a document available with this info, I would be very interested in downloading it. I have looked around and have not found one.
As an example. I currently have a RSSI from my mailbox notifier of -83. I have no idea if this is considered good enough.

[Felix]

I don't have a link for you unfortunately.
The signal goes from strong (high RSSI) to weak (low RSSI), the decrease is not linear but logarithmic and is influenced very much by the environment and obstacles.
The limit right now for the RFM69 lib is -100dBm.
At -83 I would still consider that pretty strong.
For my mailbox across the street I reduced the output power to keep it around -90RSSI. That is enough for a consistent reading.
If you're looking to increase range, you can reduce the bandwidth and bitrate. The lib has some defaults that will give you a very decent range at higher bitrate, and should cover any residential property, but those settings are not set in stone. Others have played with those to get ranges of 1-2km.