[Elecraft] K2 PLL reference oscillator range is zero -- problem SOLVED

[David Johnson]

Hello Don and the group,

  I just wanted to let everyone know what caused this problem and throw out a couple of things that I learned that might be helpful to others:

 1.  The problem was caused by the failure of U5 (LTC1451) on the RF board.  After removal of the chip it showed a permanent internal short between pin 3 (the serial interface line going to MCU pin 19) and pin 5, which is ground.  So the whole serial input line was just being permanently held at ground potential, making it impossible for the MCU pin 19 to ever rise to 5 volts.

 2. If you have K2 manual Revisions E or G (and maybe others), page 9 of Appendix A lists the part number for U5 as E600016, which is the same as the part number listed five lines down for U4.  I ordered a replacement U5 by the part number and received U4 instead (hey, at least I had a 50% chance!).  Anyway, Scott at Elecraft was very helpful and got it all straightened out for me, but for all of you really concientious types out there, change the part number for U5 in your manual on page 9 to E600030 which is the actual part number.

3.  The voltage chart on page 17 of Appendix E shows a value on pin 7 of U5 of "0 to 4".  You might want to make a note in the bottom margin of page 17 that the voltage varies from 2.0 volts to almost 3.6 volts, varying linearly across that range as the VFO knob is tuned through each successive 5 khz range.

4.  Finally, I just want to mention something that everyone who has been on this list for more than a week already knows well. We all owe Don, W3FPR a huge and endless THANK YOU for his tireless and expert assistance to all of us, week in and week out.  Don really helped me zero in on this particular problem and over the years in reading all of his posts I have learned a tremendous amount about using and understanding the elecraft radios. Thanks Don!

  -- Dave WB4JTT


--- On Tue, 1/6/09, Don Wilhelm <don at w3fpr.com> wrote:

> From: Don Wilhelm <don at w3fpr.com>
> Subject: Re: [Elecraft] K2 PLL reference oscillator range is zero
> To: wb4jtt at yahoo.com
> Cc: elecraft at mailman.qth.net
> Date: Tuesday, January 6, 2009, 10:55 PM
> Dave,
> 
> Your measurement on SCK is correct - the manual is wrong
> for that one.  If you look at it with a 'scope, you will
> see negative going pulses when the clock is active (turn the
> VFO knob to generate a few pulses.
> 
> I believe your problem is with the /DAC2CS signal which I
> just verified as 5 volts when measured as a DC signal, but
> like the SCK signal has negative going pulses when the VFO
> knob is turned.  Since your K2 /DAC2CS is at zero volts, the
> DAC is being gated for data-in at all times, so when the
> data is not being clocked to the DAC, it thinks the data
> bits in the data stream are all inactive, and the analog
> output goes to zero volts.
> 
> Your readings at U6 pins 33 and 38 are in the range of
> normal readings.  Pin 38 is active positive and is normally
> low with frequent positive going pulses, so the DC voltage
> measured can vary depending on the time constant of the
> meter used, and pin 33 is also active high, being normally
> low and having less frequent positive pulses than those on
> pin 38 - so its DC average voltage will be less than that on
> pin 38 - how much depends on your meter.
> 
> I don't know if your digital skill level was adequate
> to follow all of that explanation, but the bottom line is
> that something is not working between the microprocessor
> (Control Board U6) pin 19 and RF Board U5 pin 3.  Go back to
> CB U6 pin 19 and measure the voltage there - if the voltage
> is 5 volts there while being 0 volts at RF U5 pin 3, there
> is a break in the path - that includes the CB to RF board
> connectors - find the break and fix it (you can verify the
> fix with a continuity check on your DMM). 
> OTOH, if the voltage is zero at both places, the problem is
> *either* a bad microprocessor output or a shorted RF Board
> U5 pin 3.  The best way to determine which it is requires
> removing U5.  If the voltage goes to 5 volts with U5
> removed, the microprocessor is good and U5 is bad - but if
> it stays at zero volts, then the microprocessor needs to be
> replaced.
> 
> There is another way to check if you do not have
> de-soldering equipment and cannot easily remove U5.  Remove
> Control Board U6 from its socket and insert the ends of a
> resistor (470 ohms to 2,2k will do nicely for this check)
> between U6 pins 11 and 19 (count carefully).  Then power the
> K2 on (do not expect a display or anything else to happen),
> then measure the voltage at RF Board U5 pin 3 - if it is 5
> volts, U5 is OK, get a new K2 firmware chip - but if the
> voltage is zero at U5 pin 3, replace U5.
> 
> 73,
> Don W3FPR
> 
> Back to your problem with the U5 output
> 
> David Johnson wrote:
> > Don and the group,
> > 
> >   I have not yet solved this problem, but upon further
> troubleshooting a recurrent theme seems to be that the clock
> signal (SCK), which appears at pin 18 of MCU U6, pin 6 of
> EEPROM U7 and pin 10 of U8 on the control board, as well as
> pin 1 of DAC U5 on the RF board is constantly 5 volts when,
> according to the voltage charts, it should be 0 volts.
> > 
> >   Further, the /DAC2CS value at pin 19 of the MCU U6
> (control board), which is also at pin 3 of U5 (RF board) is
> 0 volts when it should be 5 volts.
> > 
> >   The only other anomalous voltage values on MCU U6
> are 0.1 volt on SRDIN pin 33 instead of the listed 5 volts,
> and 0.2 volt on /SR RD pin 38 instead of the listed 1.2
> volts in the chart.
> > 
> >   So my question is, what controls the SCK value on
> pin 18 of the MCU U6?  In other words, I assume that there
> is another pin on U6 that controls whether pin 18 is high or
> low.  If not, should I assume that the MCU is bad because it
> always reads high?  
> >   Many thanks to anyone willing to wade through all
> this!
> > 
> >  -- Dave WB4JTT