A full explanation will involve delving into the source code. The output voltage is set by a series of formulas starting with the Drive setting. In console.cs you can see the event handler for udPWR_ValueChanged uses the following formula for 100W setups:
========================================================================== float val = (float)udPWR.Value; double target_dbm = 10*(double)Math.Log10((double)val*1000); target_dbm -= GainByBand(CurrentBand); double target_volts = Math.Sqrt(Math.Pow(10, target_dbm*0.1)*0.05); // E = Sqrt(P * R) Audio.RadioVolume = target_volts/audio_volts1; ========================================================================== So the first thing we do is get grab the target power (udPWR.Value) and convert that into dBm (target_dbm). Then we subtract the GainByBand value for the band you are in. Then apply the gain as a voltage ratio for the target volts over the voltage for the card in use. So let's take Drive = 100. 100W = 50dBm, so target_dbm = 50.0. Let's say that we are on 20m and the gain-by-band setting is 48.0. So the target_dbm becomes 50.0 - 48.0 = 2.0dBm. Converting back to voltage we get 0.281V. Now we need to turn this into a floating point scale from [-1.0, +1.0] where the rails are the voltage limits of the card (i.e. 1.0 on a Delta 44 is 0.98V). So we take 0.281 / 0.98 = 0.287. Now that scaling would be applied to all transmit signals coming out of the DSP transmit chain which is also scaled from [-1.0, +1.0] and everything should come out just about right. Now clearly there are some non-linearities that we are taking for granted here, but in practice, this works very well for selecting power that is adjustable by band. It should be more obvious now why when you move the gain-by-band value down you get more voltage (and thus more power). The gain-by-band values represent the gain through the whole radio system (stack and PA). We gather these by setting a target wattage and working backwards. So let's say we are targeting 100W. We turn that into dB (50.0dBm) and then transmit at full power (scale = 1.0) and see what kind of power reading we get. Depending on the reading, we will adjust the gain-by-band until we are within some reasonable % of the target. At that point we know the gain for that band because we know the voltage of the soundcard and how much power we are getting out. Thus we can fairly accurately calibrate each band. Again, there are some non-linearities that we do not take into account here. These are most often seen on 160m and 12/10m. Hence the lower gain values for those bands and the less accurate power selection. Eric Wachsmann FlexRadio Systems > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > radio.biz] On Behalf Of Klaus Lohmann > Sent: Tuesday, February 06, 2007 3:33 AM > To: FlexRadio; Gerald Youngblood; 'John P Basilotto W5GI' > Subject: [Flexradio] PA calibration routine > > Eric, > a few weeks ago I asked you to provide a software version to me which > allows me to go below 38.0dB in the calibration routine table. Maybe I > have missed your answer but I found out myself that 38.0dB is already the > lower limit: at this level we have full drive from the 4-stack board. > Therefore - please - provide me with information what 38.0dB really means: > the lower the figure the higher the output. When asked by customers I > would like to feel more comfortable. > Maybe it would be a good idea to make it clear in the table itself: 0dB is > maximum and then -xxdB as appropriate on the bands. > Gerald, John, I still have problems with 2 units from customers to get > full output (100W) on 10m (13.8V at the terminal)- I get 90 Watts maximum > even after installing the MICA cap on T2. The PA cal. routine is passed > but our specs tell the customers 100W on all HF bands ... > Both units are RoHS and deliver on the higher HF-bands significantly less > than 1000mW (10m 500mW only). Will report more on that later. > Klaus _______________________________________________ FlexRadio mailing list [email protected] http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com/ FlexRadio Knowledge Base: http://kb.flex-radio.com/
