WHUUPS! Two Steps Forward and One Step Backward

OK, today was the day to get the amplifier running.  I finished wiring the 3 kW 48 VDC power supply and connected it to the 220 VAC in the shack and to the W6PQL amplifier.   After double and triple checking everything (and removing the PTT cable from the amp just to make sure I did not accidentally put it into transmit) I decided it was time to swallow a few times and hit the AC ON button.  The "pucker factor" was fairly high at this point [grin].

Well, for a couple of seconds nothing happened and I was sure I had made a big mistake.  Then I heard a relay click, fans started running in the 3 kW power supply and the amp displayed 52 VDC on the PA Voltage meter.  WHEW!!!  Time to remember to breathe now!

I relaxed for a bit and let the system just coast along.  The air coming out of the back of the 3 kW power supply began to warm a little so I might need to think about relocating that power supply.  The output of the air is blocked a little and if I am going to be running long schedules, I will need plenty of cooling.

The next step was to key the radio and see if the amp moved into Transmit.  I put the K3S on 2-M, DATA A mode, dialed the PWR control all the way to the left where it was reading -10.0dB (0.10 mW), and I stepped on the footswitch.  WHUUPS - BANG!!

There was a noise, some relay clicks and the amp displayed a red LED indicating "Load Fail."  I had been expecting nothing at all to happen except the sound of the Sequencer relays and the "Transmit" LED on the amp to come on.  BOY, was I surprised.

 After letting my heart return to a normal rhythm, I turned OFF the amp and began looking at the K3S.  Without the amp inline, when I step on the footswitch, the RF meter on the K3S jumps way up scale, not often but sometimes even to full scale.  That could explain the "Load Fail" indication on the amp by there being WAY too much drive power.  I had previously changed the input attenuator on the amp so that I could drive it with a maximum of 2-1/2 watts instead of 10-watts.  If the K3S put out a spike when it began transmitting, that may have driven the amp too far and the protection circuit shut it down.  I do also recall seeing the Forward and Reflected power indicators on the amp jump upscale briefly before the amp shut down.  I really had no plans to "test" that protection circuit but it's good to know that it is working perfectly ! ! !

I tried adjusting the drive on the K3S with a single tone on the input.  As I turned the PWR control, the RF meter would indicate full scale as I was turning it then, when I stopped turning it, the RF meter would stay at a specific level.  It almost acted like the PWR control was dirty.

I saw where there was some discussion about this online.  The comments seemed to indicate that it might be a Power Supply issue.  So, I checked the built-in meter of the K3S and it indicates a Supply Voltage of 13.6 VDC in Receive and 13.3 VDC in Transmit (PWR control at minimum.)  This is pretty much normal.

On a hunch, I increased the TX Delay in the K3S configuration from 8 ms to 14 ms and that seemed to stop the full-scale power readings when the footswitch was pressed.  At the same time I returned the input attenuator on the amp to the maximum amount of attenuation so that I could input more power without an issue.

I powered on the amp again, pressed the footswitch, and YAY!, it did NOT give me a fault!  Just the "Transmit" LED lit up.  I then put WSJT-X in Echo mode and set the PTT method to VOX so it would NOT key the transmitter by itself.  The program would be going into Echo mode transmit but until I pressed the Footswitch, the radio would not transmit.  This worked and I had about 800 watts out very briefly.  I let up on the footswitch and reduced the Audio Drive Line Input to the K3S from 17 to 2.  This time no power out was noted.  I brought it slowly up and at a level of 5 noticed that there was a spike in the Power output and a sound in the K3S speaker when it went into transmit. 

Except for the spike, there was 10 amps of PA Current being drawn by the amp but no power out shown when there was a tone being supplied except for an initial spike of 800-watts!  Without the tone, there was just 2 amps of PA Current which was probably just the bias level with no RF input.  This test was done with the footswitch putting the K3S into continuous transmit and the WSJT-X Echo mode supplying the tone for a cycle of 3 seconds ON and 3 seconds OFF.  The "spike" I observed here was not from the K3S going into transmit, it was from the tone being applied which caused the K3S to generate RF that then seemed to be "choked" down to the level of -10.0 dB (0.010 mW).

The next test I tried was just sending CW.  Without the amp on, the K3S still seemed to have a bit of a "spike" when just sending the letter "E" but sending my callsign, the RF meter only indicated 1 bar.  I set the keyer speed to 40 wpm, turned on the amp and sent a single "E" and the Forward Power indicated 800-watts out.  This was with the K3S PWR level at minimum.  I then sent my callsign and it continuously keyed at 800-watts out.  Bad news is that I was hearing RFI from one set of computer speakers - that will need to be taken care of.  But, the good news (what little there is) was that there was NO Reflected Power indicated.  Looks like Charlie, N8RR, and his famous SARK Analyzer helped me get the antennas tuned nicely!

My last step was to check the power output of the K3S on 2-M.  I pulled out a Bird Model 43 wattmeter and looked through my slugs to find what I had which would measure close to 2-M.  I had a 5W slug that covered 400-1000MHz which was close.  Using a Narda 40-watt termination that was good to 12.4 GHz, that 5W slug measured 2.8 watts output with the PWR control set to minimum.  With the K3S PWR control at -4.0 dB (0.43 mW) the slug was reading full-scale or 5-watts.  Next I used a 25W slug good for 50-125 MHz.  At the minimum PWR setting on the K3S this slug measured 4.5-watts and reached 8.5-watts with the PWR control at maximum of 1.8 dB (1.50 mW.)  That pretty much tells me that at the very lowest PWR out setting of the K3S, it is putting out enough power to drive the W6PQL amplifier to 800-watts output even with the attenuator connected.   Maybe I need a larger attenuator to be able to have more control of the drive level to the amplifier.

At this point, I'm out of ideas and ready to go "chill out" a bit.  Looks like I need to call Elecraft tomorrow afternoon and see what they can tell me.