[Elecraft] [K2] Low TX Output on Higher Bands

[Steve Kavanagh]

Well, I've been doing lots of measurements and inspections and I'm still not sure what is the matter.  
 
As suggested I rechecked the DC voltages on Q6 and got 1.03 V on the base and 0.25 V on the emitter.  The difference is 0.79 V - high, but plausible.
 
The emitter resistor R50 measures 1.4 ohms...OK.
R93 measures 809 ohms....OK.
R94 measures 83 ohms...OK.
C150 is marked 331J...OK.
Q20 seems to be working fine.
W6 to ground measures 1kohm...OK.
The ungrounded end of C212 is grounded for DC regardless of which band is selected...OK.
RF voltage is seen on Q212 only on 24 and 28 MHz, so the correct filter is selected on those bands.
8T applied to D1 is fine.
 
I measured the Q7/Q8 resistances per the manual.  With the positive lead on the base I measured around 520 ohms to emitter and collector.  With the negative lead on the base I get about 750 ohms to emitter and about 5k to the collector.  The 750 ohm number seems a bit low and might indicate something.
 
Here's my RF voltage measurement results for various requested power settings.  These are  actually the DC voltage displayed on a DMM using the RF probe circuit described in the K2  manual (though not using the Elecraft board). 
 
At W6 (in mV) 
 
MHz  1W   5W   10W 
 
3.5  3   10   16 
7       4   14   26 
10     3   11   22 
14     7   25   31 
18     5  9  9 
21     2  4  4 
24     7  8  8 
28  11   11   11 
 
The lack of change at higher power levels on the higher bands seems to agree with the fact  that the K2 output power stops increasing as the power knob is advanced. 
 
At Q5 collector (in V) 
 
MHz  1W    5W    10W 
 
3.5  .088 .175 .25 
7     .024 .070 .11 
10  .080 .22  .34 
14  .166 .36  .42 
18  .35  .51  .51 
21  .46  .72  .72 
24  .60  .64  .64 
28  .73  .76  .75 
We see much higher output and less variation with power setting on the higher bands.  This  seems consistent with a following stage having gain roll off at the higher frequencies. 
 
At Q6 collector (in V) 
 
MHz  1W  5W  10W
3.5  .89  1.5  2.2 
7     .90  1.7  2.2 
10  1.0  2.1  3.1 
14  1.1  2.4  2.8 
18  1.6  2.5  2.5 
21  2.1  3.1  3.1 
24  2.5  2.7  2.7 
28  2.9  3.1  3.1 
 
Same behaviour here.
 
At Q7/Q8 base (in V, no 1 W setting, not all frequencies measured at 10 W setting) 
 
MHz  5W          10W 
3.5  .39/.41 -/- 
7     .40/.40 .57/.62 
10  .37/.36 -/- 
14  .31/.35 .37/.42 
18  .35/.33 -/- 
21  .40/.40 .39/.37 
24  .29/.33 -/- 
28  .26/.29 .26/.28 
 
Now we see lower voltages at the higher frequencies, despite what was seen at the collector  of Q6.  This is probably consistent with the lower base impedance as frequency increases  (though feedback will modify this). 
 
At Q7/Q8 collector (in V, only some frequencies and no 1 W setting) 
 
MHz   5W          10W 
 
7    5.1/5.3 7.0/7.1 
14    5.4/5.3 6.2/6.1 
21    4.6/4.8 4.6/4.9 
28    3.6/3.8 3.2/3.4 
 
The output voltage drops with frequency and increases with power setting only at the lower  bands, which is consistent with what is seen on the power meter.
My best guess is that one of the finals (Q7 or Q8) is bad, but it isn't completely clear since the manual only gives RF voltages for 40m and I don't know how it is supposed to vary over frequency.

Any clues here for those more familiar with troubleshooting the K2?

73,
Steve VE3SMA
 
----- Original Message -----
From: Don Wilhelm 
 
Sent: Sunday, November 9, 2014 4:26 PM
Subject: Re: [Elecraft] [K2] Low TX Output on Higher Bands

Steve,

Re-check the voltage on Q6.  The emitter should be 0.6 volts *lower* 
than the base - which should put it at the manual's voltage of 0.4 
volts.  The power setting should not matter for that measurement.
Check the emitter resistor on U6 - it should be 1.5 ohms.  If this is a 
K2 below SN3000 without the A to B upgrades, R50 may be a 1/4 watt 
resistor, and it was replaced with a 1/2 watt resistor.  If R50 has been 
overheated, it is conceivable that its resistance is lower than the 1.5 
ohms.

Another thing to check is the TX buffer stage U9.  First measure the DC 
voltage at pin 3 - if Q20 is working properly, it will be at 12 volts 
(or power supply voltage) during receive, but drop to 1/2 that value 
during transmit (or TUNE).  If it does not change, replace Q20 (2N7000 
which is static sensitive).

Also check the resistance of R93 (820 ohms) and R94 (82 ohms) as well as 
the value of C150 (330pF).  Note that R94 and C150 are in series which 
makes the gain of U9 greater as the frequency is increased.

73,
Don W3FPR