Signal not getting through 3.395 MHz Crystal Filter on HW-100, then it is

Busy as usual, with CHM work yesterday on the 1401 systems, meeting this evening of the microwave experimenters group 50 MHz and UP, Alto restoration tomorrow, lunch panel discussion at Xerox PARC tomorrow, etc. 

HEATHKIT HW-100 SSB TRANSCEIVER RESTORATION AND MODIFICATION

I looked closely at the ALC voltage generating circuit underneath the final amplifier section. The 3.3M resistor is frequently cited as a cause of problems, but it measured fine when taken out of circuit. I then looked at the diodes that are part of the circuit and found the leads were all corroded, looking like they had burned or were badly eaten by something. 

I yanked one out of circuit and it tested okay, but the part is cheap enough that I bought replacements for all of them. Clearly these do not age well. I rewired the entire ALC generating circuit and checked it out, but the results are the same.

Back to tracing the signal through the transmit chain. I see the BFO (carrier oscillator) producing its appropriate frequency in Tune and CW, with that nulled out at the balanced modulator when in USB and LSB. The isolation amplifier V2, which is essentially an impedance matching circuit with no gain, yields about .5V P/P at max gain of the Mic/CW pot which seems reasonable. 

Next up, the signal will pass through a crystal filter to narrow its output to solely the 3395 KHz range with just over 2.1 KHz either side, down 6 db. I watched the signal at the input and output of the filter next. There should be a drop of about .03 V P/P across the filter. 

Ah, very interesting. The input signal is there, but zip on the output. Perhaps the crystal filter is dead. First, I will measure the actual BFO frequency with my frequency counter to be sure that I am seeing 3395 KHz. If it is off by more than 2 KHz, the signal passing through will be so attenuated that it is effectively zero. 

Typical crystal filter design

So, two possible causes of what I see. First, the crystal filter my be defective, perhaps one of the internal crystals is no longer working or has drifted too far off target frequency. Second, the input frequency is too far off the passband and is correctly blocked. If it is a bad filter, I can get a higher quality one for about $40 on eBay, from the SB line of transceivers, which is fully interchangeable. 

I can't get enough gain on the frequency running into or out of the isolation amplifier to lock up my frequency counter, thus I can't measure the frequency. I will revert to my oscilloscope again and look at the signal right off the output of transformer T1 before it goes into tube V2. This was running high enough that it should trigger my frequency counter but it didn't. 

Using the scope, I saw that in Tune mode, the BFO was running at some off frequency, while CW, USB and LSB ran at proper values. I tracked the signal at the output of the crystal filter, which did appear in CW mode. Suddenly, Tune mode was on frequency and passing through the filter!

This is enough to make me question everything I have been seeing and testing. If the BFO is not dependable, the rest of the transceiver is certain to perform improperly. It was almost time for me to head over to the microwave club dinner, a good time to put this aside and look with fresh eyes another day.