Friday, February 24, 2017

DDS working in enclosure, now debugging the transmit issues of the HW-100


This morning I tested the wiring of the DDS enclosure and for correct operation. The unit tunes up and down, the 'up' and 'down' bandswitching buttons work, and the RIT (Receive Incremental Tuning) button works properly. RIT allows me to hold the button and tune an offset so that transmit and receive occur at different frequencies, using the button to control which frequency is active.

The enclosure is just about done. I still need to mount the Heath style feet on the bottom and to put in the sheet metal screws to hold the box closed. It is operational, however.

DDS VFO wired up in its enclosure and passed all tests
I hooked the DDS and the microphone up and attempted to begin the transmitter alignment. My first try didn't seem to work, but I need to work methodically through the steps since the very first act is to set the bias; that could keep the final amplifier tubes in cutoff if not adjusted properly. 

After adjusting the bias potentiometer, the transmit tubes do draw 50ma cathode current when energized with no signal. The next step in the alignment is to switch to Tune mode and advance the Mic/CW gain until relative power is seen. It wasn't. Time to start tracing the signal through the various stages and learn where something has gone awry. 

After some work this afternoon, I had a few conclusions. First, the carrier crystals are quite close but not exactly on their 'ideal' frequency. Second, the output of the isolation amplifier is essentially zero. 

I looked at the output of the carrier oscillator, which varies based on the mode of operation. That shifts the carrier to the left or right for SSB operation, or directly on the tuned frequency for CW or Tune modes. 

For CW and Tune, the oscillator should be at 3395.4 KHz but was about 200 Hz high, an error of ,006%.  USB should be at 3396.4 KHz but was off by 20 Hz or less than .0006%. LSB should run at 3393.6 KHz but was off  by 30 Hz, also vanishingly small. I could cut a trace on the PCB and put in a trimmer capacitor for the CW crystal, but the error is too low to care about.

The output of the isolation amplifier was varying randomly from about .85 MHz to about 1.6Mhz for receive modes, then in the Tune mode it sometimes ran at about 2.3MHz and other times was off. This is not good. 

I think I should have the carrier oscillator frequency amplified here for Tune or CW, and the carrier balance control should eliminate the carrier entirely for USB or LSB. That means that output for LSB or USB will only be on the sideband away from the carrier when there is voice on the microphone during transmission. CW only passes the carrier if the code key is pressed, biasing the IF amplifier on. 

I need to back up earlier in the chain, before the isolation amplifier, where I should see the carrier oscillator signal. Next spot to look was the output of the cathode follower that drives speech signals to modulate the carrier.  

I have to admit that I performed a local modification, swapping the coupling capacitor between the speech amplifier and cathode follower/mixer to a larger capacitance to allow through a bit more bass from voice input. It shouldn't affect the carrier oscillator at all, but if I messed up the change it could be an explanation.

If the carrier is not getting through then the isolation amplifier will be boosting random coupled signals. I need to probe further back but it is slightly challenging to find scoping points. With a PCB and tightly fitted components, it takes time to find a place to hook a scope probe anywhere near the desired signal. 

Also, the high voltages require extra caution. With 800V, 350V and -130V present, I could get quite a shock. Only the final section uses 800V  but across the rest of the unit it is possible to bridge 350 and -130, almost 500V of relative potential. 
Section of schematic I am observing

If I have the signal on the cathode of the cathode follower but didn't see it on the other side of the capacitor that couples it to the next circuit, That could be because one of the diodes is shorted, since the voltage drop across working diodes would leave the AC signal observable. 

I had to redo the tests on the grid of the cathode follower, because I should see carrier on all four settings, even if the follower isn't letting that signal through. The signal that is missing comes up into the diode bridge, so that any coupling of it back to the cathode follower is coincidental. The cathode follower exists to drive the audio signal into the bridge.

The routing of the carrier oscillator signal to the diode bridge is somewhat indirect. While I know that the carrier is produced correctly for all four mode settings, I am less sure about how it gets up to the bridge and the schematic is a bit circuitous ( pardon the pun ).

No comments:

Post a Comment