Low output from 2nd crystal filter and SA602 product detector

[n7zf]

I am finishing (wishful thinking) a revision A K2 for a friend who bought the kit many years ago and did not build it before having a stroke.  I have installed all mods published on the Elecraft site, plus the Rework Eliminators.  There are no options right now (but they are built).  The filter crystals are the 14 matched ones for a K2 with SSB.

I am doing the voltage checks in the receiver.  Everything seemed fine until I reach the 2nd crystal filter output and the product detector output.  I measure about .150Vrms (expect .270Vrms) at U11 pin1.  The output from the SA602 mixer at U11 pin5 is only .060Vrms (expect .580).  I guess if the input signal to the mixer is low, then the output will be low.

Both of these voltage measurements are very subjective, because they change very dramatically if I adjust the signal generator (an HP 8640B) or the K2 VFO a minor amount.  But even doing that, I cannot get the mixer output above .250Vrms.

What I don't understand is how the rms voltage output of the 2nd crystal filter (expected .27V) can be expected to be higher than the input voltage (.15V)?  Is there some amplification going on here?

Am I off in the weeds here?  How do I troubleshoot the 2nd crystal filter?

Also, the T7 step-up ratio I measured to be 3.65 but the expected ratio is 3.88.  (I don't have the D40 and D41 limiting diodes installed.)  Is 3.88 good enough?  I have tried pushing the windings around but 3.65 is the best I can get.

My RF voltage measurements are done using a Fluke RF probe and an HP multimeter.

Bob.N7ZF

[Don Wilhelm-4]

  Bob,

The levels measured at U11 and U12 under such strong signal conditions
are a bit confusing.
Do you have AGC turned off?  If not, the gain of U12 is being reduced
dramatically.

Since you are working on an early level of the manual, check the turns
ratio used for T7 - the better ratio is 5:20 turns instead of the
earlier 4:13.  If the serial number is below 2560, I suggest that you
also change the IF crystals to the newer ones.  If the crystals are
marked "-20" they are the old ones - new ones have an "-S" suffix.  You
will have a better filter shape with the newer crystals.

I would suggest that you instead test the receiver by seeing how small a
signal that K2 can receive.  Set your HP8640 to -130 dBm and look at the
Spectrogram display.  How many dB above the receiver noise floor is the
signal pip?  If it is well above the receiver noise floor, that receiver
will hear with the best of them.

73,
Don W3FPR

On 2/1/2011 1:55 AM, n7zf wrote:

> I am finishing (wishful thinking) a revision A K2 for a friend who bought the
> kit many years ago and did not build it before having a stroke.  I have
> installed all mods published on the Elecraft site, plus the Rework
> Eliminators.  There are no options right now (but they are built).  The
> filter crystals are the 14 matched ones for a K2 with SSB.
>
> I am doing the voltage checks in the receiver.  Everything seemed fine until
> I reach the 2nd crystal filter output and the product detector output.  I
> measure about .150Vrms (expect .270Vrms) at U11 pin1.  The output from the
> SA602 mixer at U11 pin5 is only .060Vrms (expect .580).  I guess if the
> input signal to the mixer is low, then the output will be low.
>
> Both of these voltage measurements are very subjective, because they change
> very dramatically if I adjust the signal generator (an HP 8640B) or the K2
> VFO a minor amount.  But even doing that, I cannot get the mixer output
> above .250Vrms.
>
> What I don't understand is how the rms voltage output of the 2nd crystal
> filter (expected .27V) can be expected to be higher than the input voltage
> (.15V)?  Is there some amplification going on here?
>
> Am I off in the weeds here?  How do I troubleshoot the 2nd crystal filter?
>
> Also, the T7 step-up ratio I measured to be 3.65 but the expected ratio is
> 3.88.  (I don't have the D40 and D41 limiting diodes installed.)  Is 3.88
> good enough?  I have tried pushing the windings around but 3.65 is the best
> I can get.
>
> My RF voltage measurements are done using a Fluke RF probe and an HP
> multimeter.
>
> Bob.N7ZF
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[n7zf]

Don,
AGC was on,  T7 is 5:20, crystals are the -S flavor.

With a -130 dBm signal, Spectrogram shows the signal about 20 dB above the noise floor, so I guess I have nothing to worry about.  Can easily hear the signal with headphones.  This graph is with AGC on, and the AGC off chart looks about the same.  



I will go back to the original problem I was chasing:  on transmitter tests, the K2 draws about 50% more current than it should.  I am guessing that I need to work on the transformers in the PA circuit.
Thanks,
Bob N7ZF

[Don Wilhelm-4]

  Bob,

Yes, that seems quite good.

On your transmit problem, count the turns on the toroids in the Low Pass
Filter.  Count the number of times the wire passes through the center of
the core.  Many builders do not pick up on the fact that a straight wire
through the core is one turn while a full wrap around the core is two
turns - they end up with an extra turn and the Low Pass Filter's cutoff
frequency is too low.

If you have an antenna analyzer, I can tell you how to check the LPF easily.

73,
Don W3FPR

On 2/1/2011 2:09 PM, n7zf wrote:

> Don,
> AGC was on,  T7 is 5:20, crystals are the -S flavor.
>
> With a -130 dBm signal, Spectrogram shows the signal about 20 dB above the
> noise floor, so I guess I have nothing to worry about.  Can easily hear the
> signal with headphones.  This graph is with AGC on, and the AGC off chart
> looks about the same.
>
> http://elecraft.365791.n2.nabble.com/file/n5982289/minus_130_dBm_signal_at_10.1_MHz_with_agc_on.jpeg
>
> I will go back to the original problem I was chasing:  on transmitter tests,
> the K2 draws about 50% more current than it should.  I am guessing that I
> need to work on the transformers in the PA circuit.
> Thanks,
> Bob N7ZF

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[n7zf]

I have a miniVNA analyzer.

I followed the transmitter troubleshooting procedure and it warned me not to skip the receiver procedure, which is how I ended up at the 2nd crystal oscillator.  Since I have a good -130 dBm receive signal, can I assume that the low output from the 2nd crystal oscillator and product detector is not causing my high transmit current?

Bob

[Don Wilhelm-4]

  Bob,

I certainly can say that any measurements made at the 2nd crystal stage
and the product detector would have no effect at all on transmit current
draw.

How much current is it drawing?  Many K2s will develop 16 or 17 watts of
power, and if you are making your current draw judgment based on HiCur
messages when the power is cranked up to near maximum, that may not be a
problem at all, just a good performing K2 - turn CAL CUR up to 3.50.

Another thing that can cause higher than normal current draw is a
voltage that is too low.  Check using a 13.8 volt power supply.  Use the
internal voltage/current display to see how much the voltae drops during
transmit.  If it drops more than a volt, check the tightness of the
power supply leads at the supply, make certain the wire is of an
adequate gauge (I recommend #16) and make certain the wire is well
soldered to the plug.
The reason low voltage results in higher current (and not the other way
around as in many other transceivers) is due to the way the K2 controls
power - it tries to transmit at the power you set, and if the voltage is
lower, the current must be higher to maintain that power.

Also be aware that the base K2 current draw and power output must be
judged only when operating into a 50 ohm dummy load.  If you are using
an antenna, all bets are off unless you can assure me that it is 50 ohms
purely resistive at the frequency of operation.

To measure the Low Pass filters, you need to disconnect them from the PA
and terminate the input side of the LPF with a 50 (47 to 51) ohm
resistor.  If you have installed a header for the K160RX option,, it is
easy - make sure the K160RX board (or unpcb) is removed and insert the
ends of the 50 ohm resistor into the header between pins 5 and 8.  If
you followed the manual explicitly, remove one end of jumper W1.

If you do not have the header installed, lift the right end of jumper W1
and connect the resistor between the right hole for W1 and ground.  J14
pin 5 is a ground.

With the input of the LPF now terminated, connect your miniVNA (or any
other antenna analyzer type device) to the K2 antenna BNC.

Now turn the K2 on and select the 80 meter band (do not attempt
transmit).  Scan from about 3 MHz to 6 MHz with your analyzer, and you
should see a good SWR up through at least 4.5 MHz, and the SWR should
increase substantially before you get to 6 MHz.  The point where the SWR
begins to increase is the filter cutoff frequency, and that frequency
must be outside the top end of the ham band but substantially lower than
the 2nd harmonic of the lower band edge.

Switch to 40 meters and check that range - typical cutoff at 9.5 MHz
(below 2 times th low end of the 60 meter band).

switch to 20 meters to check the 30/20m LPF.  Cutoff above 15 MHz but
well below 20 MHz.
Switch to 15 meters and check the 15/17, filter - Cutoff above 22 MHz,
Switch to 10 meters and check the 10/12m filter - cutoff above 32 MHz.

If the cutoff frequency is above the ham bands, and the SWR is low
within the ham bands related to the filter, then the LPF is OK.  If not
examine the toroids and check the capacitor values.  Look on the right
side of RF Board schematic sheet 3 to identify which components are in
the LPF.

Other causes of high transmit current are problems with the PA bias and
problems with T4 - measure the DC voltage at the base of Q7 or Q8, it
should be typically near 0.6 volts, but no higher than 0.63 volts.

T4 should have the 2 turn green wire leads positioned toward Q7 and Q8,
and the white 3 turn leads should be toward the LPF area.  The "link"
wires should go straight through the hole in the side of the toroid core
and solder to the pad on the same side - it does not cross over.  In
other words there should be one wire from pad #5 to pad #6, and another
from pad#7 to pad#8.

Good luck.


73,
Don W3FPR

On 2/1/2011 4:16 PM, n7zf wrote:
> I have a miniVNA analyzer.
>
> I followed the transmitter troubleshooting procedure and it warned me not to
> skip the receiver procedure, which is how I ended up at the 2nd crystal
> oscillator.  Since I have a good -130 dBm receive signal, can I assume that
> the low output from the 2nd crystal oscillator and product detector is not
> causing my high transmit current?
>
> Bob
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[n7zf]

I am using a 50-ohm Relm dummy load.  My Astron power supply is set to 14.0 volts, which is what I remember reading somewhere in the K2 manual.  At 5 watts power, the K2 was shutting down.  

At 2.0 watts, the current is 1.5 amps and the voltage drops from 13.8 to 13.4 per the internal voltmeter.

At 5.0 watts, current is 2.1 to 2.3 amps (depending on band), voltage drops to 13.2.  The expected current per the '40-meter transmitter alignment' procedure is 1.3 to 1.6 amps, so I am about 50% too high.

At 10.0 watts, current is 2.7 amps, voltage drops to 13.1 vdc.

The low pass filters all appear good on the analyzer:
80m: min SWR is 1.04 @ 3750 kHz, max is 1.48 @ 4000 kHz
60/40m: min SWR is 1.09 from 7000 to 7080 kHz, max is 1.4 @ 60m
30/20m: max SWR is about 1.25
17/15m: 17m SWR is 1.4, 15m SWR is 1.2-1.3
12/10m: 12m SWR is 1.35, 10m SWR is 1.2-1.3
In all cases the SWR is very high at the 2nd harmonic of the lower band.

That was a very enlightening exercise, by the way, Don.  Thanks for that tip.

Visual inspection of T1 thru T4:  as far as I can see, without removing them, they look correct.  I probably could have wound T1 a little tighter.

DC measurements at the base of Q7 and Q8 are .631 and .632 volts -- right at the maximum.  

I will continue taking voltage measurements in the power amp circuit.

Bob

[n7zf]

I am still trying to find what is causing high current on transmit.

DC voltages on the PA Driver Q6: emitter is only 0.24 volts (on transmit).  Base is 1.0 VDC.

In the PA Bias circuit, the collector of Q13 is 6.6 VDC on transmit, so there is 13 mA of DC current in the collector circuit.  The expected current is 4mA, I think.  There is no current through R60, so it must be going to the finals.  The base voltage on Q7 and Q8 is .63 VDC.

Bob Willis - N7ZF