[Amps] Alpha 87A Fault 17

[Charlie Young]

Several folks have responded to my thread about the Alpha 87A fault 17.  In particular Bruce W2SE has taken an interest in this problem, even though he has no 87A and we have exchanged several email messages.  Thanks to all who responded.  Here is an update for those interested.


Summarizing, this amp came to me after a blower change by the owner.  The amp was not working prior to the change and it was giving several faults which are not known to me.  After the blower change,  the amp remained inoperable.  My friend asked for help.


I fixed a couple of minor issues in the RF output compartment which could be seen with a visual inspection.  After checking and double checking all the circuit board plug connections and doing ohmmeter tests for faults, I attempted to power the amp with the HV disconnected from the tube plates.  It blew the step start fuse but this problem was simple to correct.


The amp would power and go into the 3 minute warm up sequence.  HV was normal.  After the warm up, the amp went into standby state.  When the operate switch was pushed, the amp kicked off and generated a hard fault 1.


The issue causing fault 1 was an open wire in the harness between the microprocessor board and the output T/R module.

The 87A monitors various voltages such as Rbias, HV, filament current, etc., and kicks offline if a key monitored parameter is off or missing.  In this case, it was Rbias monitor.  The problem was not in the plug connections, it was the actual wire open.


Not only was this wire open between the cpu board and the output t/r module but it was also shorted to ground on one side of the cut wire.  I removed the AC module and the output TR module to look at the cable harness and could see no place where it was obviously pinched.  Finding nothing obvious, I replaced the wire.  The existing plugs were used; the new wire was spliced to the old wire about 1.5" behind each plug.  This repair allowed the amp to go into operate mode.

Finding the defective Rbias monitor wire was a harbinger of things to come.


An attempt to drive the amp with 10W of RF resulted in fault 17.  No IP flicker could be seen, it tripped offline so fast.

Considerable time was spent studying the amp bias circuits to confirm everything was working.  I tested the voltage drop across the pins and even the current flow through the TX pins.  The rig received when hooked to the amp, so I thought the rx pin diodes were OK. The input and output wattmeters had a difference when driven with the amp in standby.  This is where we were when the Amps Reflector message was posted.


This brings us to the present situation.  I have another wiring harness problem.  This time, it is an open wire in the operate line which activates the input and output relays.  The input and output relays are wired in series. The source voltage is a about 34VDC from a driver on the microprocessor board.  This circuit goes first to the tube deck compartment where the little input relay is located, then goes from the tube deck up to the output t/r board where the larger output relay resides.  The connecting wire between the tube deck and output t/r board is open.   I suppose having the relay coils in series provided an interlock function, so that one relay would not be open and the other closed.  Theoretically both relays would have to be in the same state.


Not only is the connecting wire between the two relays open, the associated wire that interconnects the chassis ground on both boards is shorted to the chassis, with both plugs disconnected from their respective boards.   The wire should be floating, but it is grounded.  I discovered this when considering stealing this ground wire to replace the other broken wire.  Both boards have local chassis ground.  It is a simple DC relay circuit.  The interconnecting ground wire would be redundant. Good thing I tested it first!  This grounding and opening of the wires in the harness is consistent with what happened with the Rbias monitor wire.   There is a serious problem in the wiring harness that must be found.

My next efforts will focus on this issue, even if I have to completely reverse the blower change procedures.


Here is the kicker. I think these relays were operating during my first tests, but have convinced myself  this might be wrong.  My short term memory is not the best.  I have to move my medicine bottles around when taking pills, because I will forget if a pill was just taken or not.   I take the pill, move the bottle to a completed area.  As each pill is taken, the bottle is moved to the completed area.  Sad, but this is the only foolproof way.  One of the benefits of aging.  I can live with it, just need to make adjustments.  Better than the alternative.


Why am I uncertain about the relays?  Well, the relays being open fooled the CPU and the terminal monitoring program as well.  The amp will go into operate mode on the front panel, the operate led will light, and the monitor program shows the amp moves to operate from standby.   If you key the amp with a footswitch, the pin diode bias will change state from receive to transmit.  The tube will draw idle current.  The input and output wattmeters are in the circuit all the time.  On the input wattmeter, a tap feeds RF into the frequency counter, which can detect RF drive even if the input and output relays remain open.  If the amp is on a different band from the transmit RF, the amp band segment will change and the presets will move;  the input/output relays are open during the retuning process anyway.  Only after the amp resets the band info will the relays close.


If the relays were open and the amp were driven with RF, the thing should immediately fault.   There is at least an even chance this was the problem after we fixed the fault 1/open Rbias monitor line issue.   I have convinced myself it is at least possible these relays were not closing but I just assumed they were; the amp and computer said it was in operate mode.  When I drove it with RF it looked like it was in operate mode.


If I missed this relay problem, it was a fundamental oversight.  On a conventional amp, the relays would be the first thing I check, after confirming filament voltage and HV.  The time was not wasted though, because the process of troubleshooting was educational about pin diodes and the amp control circuits in general.


Now we will check the wiring harness again, run a new relay wire, and see just how bad my short term memory is!  I  actually hope the amp works after this wire  repair. If it is still inoperable, we will keep digging!


Regarding this wire problem, I can envision a scenario where this open operate wire became grounded  by whatever cut it.  If this occurred, the circuit to ground for the input relay would have been completed, and it would have energized with the full 34VDC across it.  The output relay would have remained open.  The amp if driven could have made power into an open circuit on the output side.  Fault time again.   We are not going to borrow trouble!


73 Charlie N8RR