Sunday, July 21, 2013

Repairing the Alpha 87a Amplifier

Recently I mentioned on the WVDXA Reflector that I was running barefoot because my Alpha 87a was on the sick list.  I was blown away by the many generous offers of help from WVDXA members.  People offered to lend me amplifiers from all corners of the state.  Those kind offers of help told me in no uncertain terms that the WVDXA as a whole sincerely wanted me to get out of the QRP category - and quickly!  That motivated me to move the repair of my Alpha 87a to the top of my list.

Click on any photo to see it larger.
Screws to Be Removed

I am embarrassed to say that my amp went out more than 20 MONTHS ago!  The issue was that as soon as the PTT line was keyed, the amp faulted and indicated the PIN back bias voltage was not at the minimum required level.  This is known as a Fault 1 problem.

When it got sick, Charlie, N8RR, took the amp in for repair.  He was able to reproduce the fault every time he keyed the amp.  As soon as he started measuring the R Bias voltage, he noticed a loose connection which he tightened up.  Once that was done, the Fault 1 problem never reappeared.  He tried the amp out at his QTH and it made full power with no trouble.  We both were relieved and thought Charlie had made a simple fix that solved the problem.

However, once the amp was back at my QTH, Mr. Murphy saw to it that the Fault 1 problem came back.  I took apart the R Bias connection that Charlie had found to be loose, thoroughly cleaned it and "firmly" tightened it back up - but the Fault 1 issue was still there.
T/R Module under Screening
At that time, RF Concepts (Alpha Amplifiers) said that if I shipped it to them for repair, it would sit in line for a minimum of 90 days before they could even begin the repair.  I figured I could probably fix it sooner than that so I just ordered all the parts that were listed in Alpha's "Troubleshooting Guide" for the Fault 1 problem as being possible failure points.  That was 4 PIN diodes, 1 power diode and 1 resistor.  Those six parts in their individual packing bags, failed to weigh even one ounce.  Still, it cost $65 to purchase them!

Once the parts had been received, I looked carefully at the procedure to replace everything and was dismayed.  5 of the 6 parts went on the T/R Module PC board which is a real bear to remove.  I have been in that compartment twice before over the 22+ years I have owned this amp and did not relish the job.  30 screws have to be removed just to get the cover off the amp.  Then another 15 or 20 screws, nuts, flat washers, and lock washers need to be removed.  Then, wires need to be unsoldered just to get at the PC board where the parts are to be replaced.   Finally, two smaller boards have to be removed along with the T/R Module.  I knew that I needed at least a couple of days where I would be doing NOTHING but working on this amp in order to have any chance of succeeding.  If I started this project and then let it sit while I tackled something else, I would end up with a box of parts and pieces that I could never get back together.  Plus, most of my free time in 2012 was taken up with the NH8S Swains Island DX'pedition.

I was not without an amp for the full 20 months as Charlie, N8RR, graciously lent me an AL-82 for about 6 months.  He would have let me use it longer but I knew that I had to return it or I would not get off my duff and repair the 87a!  Thanks Charlie!

T/R Module in Sub-Chassis
But, once I was spurred into putting this project on the front burner by the well wishes of the WVDXA, I decided to set aside this weekend to fix the Alpha.  I started bright and early on Friday and worked on the project until about 9 p.m.  I did not work constantly but took several breaks so that I did not get over-stressed.  Trying to get that T/R Module out of the metal sub-chassis it lives in nearly gave me hives!  At one point I sat and stared at the box for nearly an hour trying to figure out the best way to tackle that situation.  In the photo at the left you can see how the two smaller boards and the T/R Module are all shoe-horned into the metal sub-chassis.  I had instructions from Alpha on getting the sub-chassis out of the Alpha but no instructions on how to get the PC board out of the metal box.  You can also see the RF Input cable coming in from the bottom and the pin where it is soldered on the bottom left of the PC board.  You can also see the black "mistake mark" on the white wire above it.  Getting a small soldering iron into that space to unsolder that wire from that large pin was a challenge.

I was finally successful in getting the board out of the sub-chassis without breaking anything and with only one small burn to a wire (made by a slip of the soldering iron.)  Once it was free I took off a couple of hours to decompress!

When I came back to the project, it was not hard to remove the PIN diodes.  However, I found that two of them had resistors soldered directly (and tightly) across them.  I had not ordered replacement resistors so I needed to carefully unsolder those and salvage them for the new diodes.  That was tedious but again I was successful.  I was aided in removing the parts from the PC board by the fact that I had acquired (as a Christmas present several years ago) a neat desoldering tool.  This is basically a soldering iron which is hollow and has a vacuum pump attached to it.  Heat the joint with the iron and when the solder flows, hit the switch and suck the solder right out of the hole!  Amazing!

Diode/Resistor String I Built
I went to bed on Friday night after replacing the four PIN diodes (two with salvaged resistors across them - see photo at right) and one replacement resistor.  This left only one power diode replacement on the HV PC Board and putting everything back together.  Saturday morning I put the T/R Module back in the sub-chassis and it went in smooth as butter!  I was able to re-solder the wires without burning anything else and managed to put all the screws, nuts, bolts, flat washers and lock washers back with nothing left over.

Now I had to figure out how to remove the HV board.  Luckily I found several photos on the Internet (HERE) posted by a guy who had done that so my worries over how to remove that board were moot.  I pulled the HV board out in under 5 minutes (record time compared to the stress-filled adventure of taking out the T/R Module.)  Replacing the diode was done in about 10 minutes and the HV board went back without a hitch.

I took the opportunity during the repair to remove what dust I could from inside the amp.  In working on the T/R Module sub-chassis, I removed the 3CX800A7 tubes and gave them a thorough cleaning as well.  I used compressed air liberally to blow out what dust I could and used anhydrous alcohol to clean most things.  Once the repairs were completed, I spent about 15 minutes looking over everything to make sure I had not screwed up anything.  During that check I did find one cable I had forgotten to re-attach!  Whew!  Dodged a bullet there.

Trying to Get T/R Module OUT
I decided not to try and test anything but I just went ahead and completely re-assembled the amplifier including the 30 cabinet screws.  Once it was all back together, I lifted the monster up to my shelf and hooked up the RF Input and Output coaxes and the PTT Relay line.  I was ready to plug it in to the 220 VAC jack but now I had doubts.  I took a break and struggled with whether I had done all I could do and I wondered if being off for 20 months would have any negative effects on the amp.  But, I decided to give it a go anyway and I was within 2 minutes of plugging in the AC cable when the storm knocked out our power!  Was this an omen?  A sign?  A warning?  Who knew?

So, Evelyn and I left to run some errands and ended up at Pizza Hut in Milton (since Teays Valley, Scott Depot and Hurricane were all without power.)  We got home about 9:30 p.m. to find the power was back and after eating a bite I plugged in the amp and said a small prayer before pushing the switch. 

Lo and behold the amp fired up, went through its boot up sequence and I marveled at how nice those LED's looked after the amp was dark for close to 2 years.  When the time delay completed, and the amp switched to READY, there was a loud noise and the amp went dark.  RATS!  Murphy struck again. 

T/R Module Removed
It turned out that this Fault was a Fault 24 (System Voltage Fault) and I presumed that was due to the amp being off for so long and the capacitors had lost all their charge.  I bit my lip and powered it on again and this time it stayed on.  Hurray!

Now to see if I had managed to fix the Fault 1 issue.  I put the amp in operate and closed the PTT line with no drive applied.  20 months ago this would result in the amp Faulting with a Fault 1 code.  But, this time, NO FAULT!  Yippee!

Now to try the RF power test.  I had been planning to test on 30-M but moved the rig to 17-M instead.  That's where I made my first mistake.  I had turned down the power level of the K-3 to 5 watts but that was on 30-M.  On 17-M, the power level was still set to 109-watts!  As soon as I hit the PTT, the amp faulted with a Fault 14 code (Excessive load reflected power or RF voltage.)  Should have known that would happen.  65-watts of drive should give 1,500-watts out so 109-watts of drive was just a bit too much.

Technician At Work!
Finally, after correcting all my "cockpit errors", I managed to successfully operate the amp at over 400-watts output with 40-watts of drive.  The SWR was 3.22:1 on my 17-M dipole and the reflected power was about 130-watts so this was as high as I dared go until the antenna can be repaired.  I then moved to 30-M and with 8-watts drive achieved 196.5-watts output.

This testing was a bit limited, I was not able to test power output much over 400-watts, but I'm cautiously optimistic that once I correct the SWR problem on the 17-M antenna, I will be back in business.

In the photo of me working on the amp, note that I have added an LED headlamp just above my Optivisor.  That helps really put the light right on the spot where I need it!  And, I have made no mention of the screws I took out that did not need to be removed, nor the 8 or 9 items that I dropped and had to search for on the floor.  That's all forgotten now!

The next step is to hook up the Palstar AT-Auto tuner and get the 160-M Inverted-L back up.  That should let me operate at or near full power on most bands like I have done before.  This is, of course, a temporary situation.  My efforts now need to be directed to fixing the SteppIR so I can again have a "real" antenna on every band.

So, I give a great big "THANK YOU" to the WVDXA members for spurring me to finish this project.  The job was not as big as I had envisioned and I did not "kill" the amp while trying to repair it.   But, don't ask me to dig back into that T/R Module sub-chassis anytime in the next DECADE!