Examine the results, and set the attenuators such that on the band with the lowest output
power the attenuator is set to 0dB, and the other band’s attenuator setting give a matching
output power. In my case this would be:
Band:
160
80
40
30
20
17
15
12
10
Atten:
!
1
!
2
!
2
!
1
!
2
0
!
1
!
1
0
Atten:
!
2
!
2
!
3
!
2
!
2
!
1
!
2
!
2
!
1
Note that the second row shows an alternative setting using slightly lower overall gain, but
achieving the same result. As you select the various bands, the attenuator setting will be
preserved and saved to the EEPROM when the transceiver is powered down, thus you only
really need to “calibrate” the transmit chain once. If you wish to obtain a record of the
attenuation settings, then use the serial test facility to dump the current values.
Note that if you restore the factory default settings, as well as inserting the measurement
system’s calibration values, you will have to re-select the various attenuation settings for each
band.
It was whilst making these adjustments that I discovered that initially my 40m output was
about 3dB lower than the rest. A careful check with my sampling oscilloscope showed that
the culprit was the 4Mhz – 8MHz input filter formed with C25, L15, C26, L16, C27, C28,
L17, and C29. A close examination showed that one side of C26 had not been soldered.
When this was corrected, the measured filter insertion loss was:
Band (Metres)
Filter (MHz)
Input (mV P-P)
Output (mV P-P)
Attenuation dB
160
1 – 2
808
648
!
1.92
80
2 – 4
808
688
!
1.40
40
4 – 8
752
744
!
0.09
30
8 – 15
920
680
!
2.63
20
8 – 15
744
752
+0.09
17
15 – 30
904
656
!
2.79
15
15 – 30
792
688
!
1.22
12
15 – 30
752
680
!
0.87
10
15 – 30
784
670
!
1.36
The output power was set to about 1.8W on 160m just to keep the signal levels well within
the derating range of both the oscilloscope input and the x10 probe used.
Adrian Ryan – 5B4AIY
47