QRZ.COM
ad: wmr-1
ad: L-HROutlet
ad: l-rl
ad: L-rfparts
ad: l-assoc
ad: l-gcopper
ad: Subscribe
KG5NII USA flag USA

Login is required for additional detail.

Email: Login required to view

Ham Member Lookups: 775

 

I am a new ham. Wet and green behind the ears. I took my element 2 and 3 tests on 30-Apr-2016 and got on the air 06-May-2016.  I was a novice back in the early 1970's when I was 12 or 13 years old. Before I could get on the air had to learn CW.  When I was 10 years old I built a 3 tube Autodyne receiver from a 1940 Radio Handbook.  I didn't have anyone to help me, and spent almost an entire summer trying to get it working.  I finally did and learned a lot.  I taught myself CW by listening to commercial ship to shore stations as well as amateur radio operators in the CW bands.   That old receiver served me well until my dad bought me a Heathkit HR-10B receiver on my 12th birthday.  I took my time with it and made sure every component was perfect in the kit.  I finally passed my novice license later that fall.  My father then bought me a used Heathkit DX-60B - the matching transmitter to the HR-10B receiver.  A homebrew ground plane for 15 meters completed my station.  I had just one crystal: 21.126 Mhz if memory serves me right.  I had to wait almost 9 weeks for my license to arrive.  It was pure torture.  The passage of time has changed so many things.

 

Last 15 DX or Special Entries

 

 

30 M QRP Homebrew Rig

 

The Black Beauty is a 30 M QRP rig that features a AD9850 DDS chip, and a Arduino Mega 2560 microcontroller. The rig outputs a full watt, has a built in keyer and features a unique joystick which allows for the easy adjustment of tuning step rate as well as a way to navigate up/down the band by moving the joystick up to move up in frequency or down to move it down. Additionally it also is equipped with a quadrature encoder and large tuning knob for traditional frequency change. The rig also has a customized tone decoder using the LM567 tone decoder chip which is used as a CW tuning aid. This is _NOT_ a DC receiver but is a superhetrodyne with IF crystal filtering as well as AGC.  I could have done a better job for the front panel,  but it's good enough for me and thats all that counts. The receiver is extremely quiet and sensitive and I also installed a 250 Hz cw audio filter which has absolutely _NO_ ringing in the audio.  The 'S' in the upper left of the LCD display is the beginnings of a s meter.  I still need to interface the receivers AGC to one of the analog input ports of the controller.  I will probably use an opamp as a reference source.   I am still  working out all the little kinks and bugs but am making pretty good progress.  I have been working on the software which is written in C/C++  and recently I have begun to profile the code and clean it up.   I plan on adding a RIT but quite frankly I have not given  much thought as to how I will accomplish this.   The main tuning knob is a dream to use.  In fact I like it better than the one on my Yaesu FTDX-1200!   As time moves on, I will probably be adding more features such as digital pots for volume control, keyer speed control, rf gain and RIT.    The front panel is made out of Lexan and is not metallic.  I simply painted the INSIDE of the panel flat black.   The orange LED in the top center of the LCD is the cw tuning indicator.  It will start flashing to the dots and dashes of the received cw signal when the cw tone is about 715 hz or so.  It's adjustable.  I will be building a little amplifer for it.  I'd like to get about 50 w output power. 

 

My first real VOM

 

By the summer of 1974 I was really getting into electronics and desperately needed a good VOM. Anything I wanted in those days I had to work for so I worked all summer long to save up and buy a good VOM. In November of '74 despite the catalogue saying differently, Radio Shack had in stock their 27 Range FET VOM around Thanksgiving. It was $59.95 which was an absolute HUGE amount of money for a kid to save up in that day and age. But I had managed to save enough for the VOM plus the sales tax and so my mother drove me to Radio Shack to purchase it. The salesman told me I needed batteries, and when I pointed out the catalogue said it had batteries included, he assured me it was a mistake that batteries were not included. Well I only had the money for the VOM plus the sales tax. So I purchased it but was totally bummed out about not being able to use it. On the drive home my mother detected something was wrong and she being a mom dug it out of me. So I told her. We drove on a ways and then suddenly my mother pulls over and turns around and starts heading back to Radio Shack. She acknowledged  I had been saving up for a long time to buy that instrument and it just wasn't right to have it in hand and not be able to use it because it had no batteries. So she informs me she is going to buy me the batteries and some spares and "thats all there is to it and I'm not gonna hear another word about it!"

 

And so I got the batteries and the spares! I get the unit home and open it up and behold the batteries were included!  Telling my mother about this, she calls Radio Shack and really gives the manager an earfull over the phone about how they were lying SOB's that tried to deceive her son!  My mother wasn't to be messed with when she was angered and bless her heart she could curse like a sailor and drive a preacher to cuss if she was motivated to do so!  And in this case she was motivated!  So she ended up  giving that Radio Shack manager a lesson or two about how being a lying, cheating, thieving SOB of a salesman can be bad for business because she's gonna tell everybody she knows and anyone who would care to listen about  how you folks deceived my son! I can imagine that poor manager must have been horrified at the thought of my mother picketing that store with signs of God knows what kind of verbage on them!  She ended up getting a full refund for the batteries with Radio Shack giving them to me as a courtesy.  You just didn't mess with my momma!

 

I still have and use this meter to this day.  It's a little worn around the edges, been dropped a number of times, has various broken plastic pieces mended together with super glue and transparent Scotch tape, and is missing the handle. But it works great!  It has outlasted every DMM  I have ever bought and I have had quite a few of them!  42 years old!

.

My first receiver

 

This is the schematic of the first receiver I built when I was about 10 years old.  Mine had a third tube which was a voltage regulator used to replace the batteries.  I ran a higher plate voltage if memory serves me right.  Pretty simple, but for a 10 year old I had to learn a lot of different skills in order to make it work. I had to learn how to read schematics, how to identify the components, how to solder them, how to drill the holes for the sockets etc.   Next, money was tight and I had to earn the money by doing odd jobs, mowing yards etc.  Lucky for me though there were plenty of old television sets that could be had that had all kinds of parts in them.  Sockets, tubes, resistors,capacitors.  The variable capacitors were about the only thing I had to buy.  My antenna was a hobo antenna of any kind of wire I could find: zipcord, bell wire, magnet wire. It was the most pathetic looking thing.   So I spent most of the summer of 1968 figuring it all all out and I finally got it working. It had a beautiful white front panel made out of masonite with a brand new vernier drive from Radio Shack attached to the bandspread variable capacitor. I fed the output of the receiver into my little solid state stereo amplifier I had saved up and bought from  a neighbor since I didn't have any headphones.  It was with this receiver that I  learned CW by listening to coastal stations send and receive traffic  from ships in CW.

 

New antenna analyzer

Antennas are hard! There is a lot of work (pain) involved in getting them tuned correctly and without an antenna analyzer to see what is going on then one may as well just put on a blindfold and spin a roulette wheel.  With this in mind and given the fact I have absolutely no interest in VHF, I after careful consideration purchased a RigExpert AA-30.  Works to 30 Mhz.

 

 

I use the Linux operating system exclusively and it is good news to report that the RigExpert Antscope software runs under Linux using WINE. No fancy sofware installation.  No fuss no hassle.  Just wine ./antscope.exe makes the magic happen.  Sweetness.  But you do not even need to use Antscope at all.  Or even read the manual for that matter - at least for the common functions one will use the most.   I used the AA-30 to analyze and adjust my 40M dipole.  Ended up trimming 24 inches off of it.  As for my 20M ground mounted vertical monobander (Marconi) I found it had a resonate frequency of around 13.85 Mhz.  But, the antenna's  bandwidth is so wide that I still had reasonable SWR across the entire 20M band.   By fine tuning it I have pretty much a  flat 1.2:1 across the entire 20M band.  As Austin Powers might say - Groovy Baby!  And here is something else groovy for you hip ham-cats out there:

 

 

Whats that say? Made in Ukraine. None of this Made in China baloney like almost every product you can buy is here in the USA.  If its made in China its crap. Don't even get me started on how I think US companies are ecomonic terrorists whose greed has turned America that was once a heavy industralized nation into a nation of fat, overweight couch potatoes addicted to their stupid Internet connected wireless devices.   Unfortunately, we do not see many labels that say Made in USA on them but it is refreshing to see one that does not say Made in China.

 

 

2016 Field Day

When I was a novice the opportunity never presented itself to participate in Field Day. I had always wanted to see what it was like though.  Anyone who has ever been in the Texas Gulf Coast area will be able to relate to the fact it is so hot and humid here during the summer months.  There was just no way I could endure this heat by doing field day the way I had always envisoned it should be  done:  from a campsite using emergency power.   So instead I opted to  do it from home as class E:   home station running emergency power.  I hooked up my lawn mower battery to my FTdx-1200 and reduced its power output to about 25 watts. By doing it this way the battery lasted the entire time without a recharge!  I was in no big hurry.  I only got on the air when I felt like it.  I had decided to just use  CW only as I just didn't want to fiddle with using the digital modes and I wouldn't even consider using phone.  Anyway off I went starting out on 20M, switching to 40M and then ending things on 15 M.   I had not made any contacts on 15M  since my novice days.  When field day first started it was as if I had completely forgotten CW!  I just couldnt seem to copy  class and the section info for some reason.  It took a number of hours of struggling in this way before I finally got myself accustomed to it so I got off to a slow start.  I took a couple of naps in between as well.  I really wasn't in any hurry and I really didn't care how many qso's I would end up making. If I got tired I would stop and take a break and if I felt like doing some more I would! All in all it was a lot of fun doing it this way.  Let me say that my Yaesu FTdx-1200 has an absolute fabulous receiver with superior filtering.  Copying weak signals in amongst stronger ones on very adjacent frequencies just wasn't a issue.  It performed just way beyond  my wildest expectations and I now have a new level of respect for this excellent rig.  Anyhow, as I said I was in no big hurry and just coasted along at my own pace.  My end results were something like this:

 

  40m   105  Qso's    using 40M 1/2wave dipole @25 watts

  20m   129  Qso's   using 20M ground mounted vertical @25 watts

  15m     15  Qso's    using 40M 1/2wave dipole @25 watts


Call Used: KG5NII

ARRL/RAC Section: STX     Class: 1E
Participants: 1    
Power Source(s): Battery
Power Multiplier: 2X
Bonus Points:
  100% Emergency power                           100
  W1AW Field Day Message                        100
  Submitted via the Web                               50
Total Bonus Points                                     250
Score Summary:
                        CW  Digital  Phone  Total
 Total QSOs    249      0         0
 Total Points   498      0          0         498  

 

Claimed Score = 996

So there it is.  Not a very high score, Jimmy crack corn I don't care.  It was fun and I really enjoyed myself.

 

 

20 Meter Vertical Project

 

I wanted to check out 20 meters and after considering my options a ground mounted vertical was deemed to be the best route to take.  The cheapest commercial 20 meter monobander I could find was in the neighborhood of around $150.00.  Commercial multi band verticals were from about $189.00 on up.  If I opted to buy a commercial antenna I could learn a lot  but I could learn more by building my own and maybe save a few bucks.  One has to keep in mind the following when looking to buy a commercial antenna:

 

  1. The initial cost of the antenna itself.
  2. Cost of the feed line from your shack to the antenna
  3. The antenna's ground system which is typically a system of radials
  4. The height of the antenna and whether or not it is engineered as self supporting or needing to be guyed.  If guyed, the cost of the guy support system.
  5. Ground mounted or put up into the air.  There is a cost associated with both.
  6. Maintenance cost and the ease of maintenance.

 

Given the cost of commercial antennas I could only afford the low end 5 band verticals in the $189 range.  The problem was given the cost issues above, the cost of the coax and the ground radials. Addressing the maintenance issue, the cost for adding a tilt base to the mix was in the order of $75.00 to $99.00.  I could have built a tilting base myself but quite frankly it may not have been engineered as well as some of the commercial units but it would have been adequate enough.  It was just a question of whether it was cost effective to attempt to build one. 

 

Upon evaluating my options and looking at a homebrew solution I noted that several companies sold telescoping aluminum tubing:  One was Texas Towers and the other was DX Engineering.   Now living in Texas I am all for supporting the local merchants but there were two issues with Texas Towers.  1 was the amount of money they charged for shipping was right on par with the actual cost of the tubing.  The other was the sales tax.  So the two taken together  exceeded the cost of the components I was actually buying.  I figured at this point that it was game over.  I considered driving to Plano, TX which from my QTH is about 250 miles - A 3.5 hour drive one way.  Nope.  Cheaper to pay the shipping and the tax!  So unfortunately I had to find a better way.  To be clear on the matter though, the folks at Texas Towers seem to be honest and decent folks and if I  wasn't on a strict budget I could justify using them if I bought a larger quantity of tubing for future projects.  But my budget wouldn't allow that as an option.

 

 

I checked out DX Engineering's site and they had the tubing.  A larger selection even.  Historically when looking for something I have always felt that DX Engineering was more expensive then most and perhaps they are on some items.  But the tubing, clamps, coax prices were reasonable, and since they were out of  state, then I automatically save an additional 8.5% I would have paid in sales tax.  So it came down to shipping cost.  I ordered the following for my  vertical:

 

  • DXE-AT1209 Aluminum Tubing, 6063-T832, 6 ft., 1.000 in. OD, 0.058 in.Wall, Slit End
  • DXE-AT1208 Aluminum Tubing, 6063-T832, 6 ft., 0.875 in. OD, 0.058 in.Wall, Slit End
  • DXE-AT1207 Aluminum Tubing, 6063-T832, 6 ft., 0.750 in. OD, 0.058 in.Wall, Slit End
  • DXE-AT1206 Aluminum Tubing, 6063-T832, 6 ft., 0.625 in. OD, 0.058 in.Wall, Slit End
  • DXE-AT1205 Aluminum Tubing, 6063-T832, 6 ft., 0.500 in. OD, 0.058 in.Wall, Slit End
  • DXE-SO239BKT-1 SO-239 Mounting Bracket
  • DXE-ECL-040 Element Clamp, Stainless Steel, 5/16 in. Width, for 0.625 in. Tube O.D.
  • DXE-ECL-060 Element Clamp, Stainless Steel,1/2 in. Width, for 0.750 in. To 0.875 in. Tube O.D.,QTY 2
  • DXE-ECL-10SS Element Clamp, Stainless Steel, Natural, 1/2 in. Width, for 1.000 in. Tube O.D.
  • DXE-8XDX075 RG-8X, 16 AWG Copper, Type II-A Non-Contaminating PVC Jacket, PL-259s, 75 ft.
  • DXE-P8A Anti-Oxidant, Penetrox A, 8 oz.

 

The total was about $112.00 with the coax being the largest expense.  Their shipping was a very reasonable $15.45.  I placed the order on a early Wednesday morning thinking it would take a week to arrive, but to my surprise Federal Express delivered my parts that Saturday! Not bad!  Not bad at all!  Now the  75 ft of coax was $45.00.  I considered buying something better but the RG-8X has only 1.4 db loss at 30Mhz so the loss at 14Mhz is even less.   And I'm using every bit of that 75 foot length.  Other expenses from Home Depot:

 

  • 2 ft. of  1 in. schedule 40 PVC
  • 10 ft. of 1.25 in schedule 40 PVC cut into two 5 ft. sections.
  • Stainless steel bolts, nuts, washers and lock washers
  • 500 ft. roll THHN 14 awg solid copper black
  • 100 ft. 14 awg solid galvanized steel wire (staples to hold the radials down)
  • Extra Heavy Duty All Weather Duct Tape

 

Construction

The problem was how to mount the vertical in the ground.  With all the aluminum tubing attached the entire assembly weighs less than 5 lbs.  With that in mind this is what I came up with.  A one inch piece of schedule 40 PVC pipe has an ID of 1.033 in and a OD of 1.315 in.  The 1 in.  piece of aluminum pipe will fit perfectly into this.  A 1.25 in. schedule 40 PVC pipe has an ID of 1.363 in. So a piece of 1 in. PVC pipe will fit perfectly into this.  I took a 5 ft. section of 1.25 in PVC and drove it 3 ft into the ground. The 1 in. PVC is bolted to the 1 in. 6 foot piece of aluminum tubing such that half is over lapping the tubing.  This 2 ft. section is then put inside the 1.25 inch PVC that is driven in the ground. I had an old  4 or 5 foot piece of 1 inch copper pipe that I drove in next to the 1.25 in. PVC to give it some extra support.  I used extra strength duct tape to temporary attach the 1.25 in. PVC to the copper pipe.  I didn't have any saddle clamps on hand so the duct tape still worked out well. But I do plan on adding the saddle clamp.

 

 

As can be seen in the above photo, I decided to paint the radial bar at the last moment.   Actually I painted everything with a connection with primer as a after thought.  Clearly I wasn't trying to impress anyone.  The reason for my haste was it was extremely hot and humid this particular day and in the weeks previous our area had experienced a very unusual amount of rain.  In fact 3 weeks earlier I was flooded out.  So the earth was water saturated and this made a prime environment for mosquitoes gone wild. Used mosquito repellant but it would just sweat off in minutes and between the heat, humidity and the mosquitoes I wasn't in any mood for perfection!  I still need to dig a trench to bury the coax, but given the mosquitoes gone wild experience earlier I am in no hurry to do this until things dry up some.  Of course then it will be more difficult to dig the trench!

 

 

Using the formula for a 1/4 wave antenna, 234/14.175MHz gave me 16.5 ft.   I adjusted the telescoping tubing as follows:

 

Except for the 6 foot piece of 1 in. tubing which is the base end of the antenna, 3 tubes were adjusted to 2.5 ft.  The last section was adjusted to 3 ft.  (6+3*2.5+3).  The overlapping telescoping sections makes for a very rigid radiator.  If I wanted to get on 30 Meters I could just adjust the lengths of the tubes to  23 ft but I'm quite lazy and I doubt I would go to this trouble.

 

For the moment I laid out only 20 radials and I have 10 more to lay out.  See the reference to mosquitoes gone wild above for an explanation as to why I stopped at 20!  Presumably radial length for a 1/4 wave vertical is supposed to be calculated by: Lft=(234/FMhz)*CoaxVelocity so given a velocity of 87% for RG8x and a center frequency of 14.175Mhz: (234/14.175)*.87 = 14.4 feet. So I ignored this and made them 16.5 feet.  I have no easy way to determine which length is optimal. I feel like this is close enough.  More precision would yield diminishing returns for the extra effort that I'm too lazy to bother with.

 

 

At first I thought I would take my gas edger and cut some grooves in the dirt for the radials but this proved to be difficult to do in a straight line so after doing 8 or so I stopped and just laid them out on the ground.  I used some 14 awg solid galvanized steel wire cut into 10 inch lengths bent into the shape of a 'U' to secure the radial wire to the ground.  3 per radial but I plan on adding 2 more per radial .  These 'staples' worked out pretty good only because the ground was moist.  Had the ground been dry I think they would not have worked. 

 

 

A strange thing happened when testing.  The first day I ran out of time as it got dark before I could lay the radials down.  So I  ran some braid from the  so-239 bracket to the 4 foot copper pipe that I had  hammered in next to the PVC pipe.   Testing it for the first time it had a 1.1 to 1 SWR across the entire 20 Meter band.  Sweet!   And it got fairly decent signal reports.  The next day I laid down the radials.  The SWR  was about 1.5 - 1.6 to 1 across the entire 20 Meter band.   O.K..  I expected the  feed point impedance to change when adding the radials.  Strange thing though was the antenna would not tune until I disconnected the 4 foot copper pipe off the system. Not sure what is up with that except the distance from the pipe to the shield of the coax is much shorter than the radial connection bar is.  Testing it with the radials laid out resulted in some pretty good reports but the jury is still on its performance.  I might just leave well enough alone.  I figure I could fiddle with it and just make things worse. Time and time again I end up tweaking and twiddling things until they get hopelessly screwed up! Not this time! I've learned my lesson.  It works.  It seems to work pretty good.  LEAVE WELL ENOUGH ALONE!  Except that I can't.  At least not completely.  I need to add some caulking to the area where the 1 in. aluminum tube enters the 1 in. PVC in order to keep it from retaining water.   Secondly I need to come up with a better method of connecting the radial buss to the SO-239 bracket.

 

What was the final cost?  $128.00 at DX Engineering and another $60 at Home Depot.  Add  a huge amount of itching, scratching, cursing and about a pint of my blood sacrificed to the blood sucking mosquitoes gone nuts.  About  $190.00 all together.  I could have got a 5 band for that but I still would have had to buy 75 feet of coax and the wire for the radials.  I believe I have a good baseline to modify into a multi band if I wanted.  Overall I am happy thus far.  The design seems to be quite robust.  I can add another 6 foot tube and make it a 1/2 wave but I would probably need to have an external ant tuner for that and I'm too lazy anyway.

 

Good news!  The antenna tunes up on 17M and 12M using my FTdx-1200's built-in antenna tuner and Bob's your uncle. It seems to get out pretty good too.   Check my recent logbook activity at the beginning of this page. 

 

 

My 50 amp power supply rebuild project

The Blak Boxen:

 

The front panel is difficult to photograph due to the glare.  C'est la vie.  The front panel has  Lexan polycarbonate sheet  painted flat black on the back and overlayed over the original aluminum front panel.  It was necessary to do this in order to hide the three rectangular cutouts for the power switch and the front panel meters showing voltage and current.  I chose Lexan instead of acrylic sheet (also referred to as Plexiglass) because acrylic yellows with age and is somewhat easy to scratch.  Lexan polycarbonate is scratch resistant, shatter resistant and is about 250 times stronger than glass.  In small sizes it is somewhat cheap and can be purchased from Home Depot or Lowes.  The photograph does not do the front panel justice.  I should probably put on some bold white lettering.  That would make it look more professional I think.  The top cover was spray painted  with flat black enamel and then a liberal coat of Krylon matte clear enamel was applied to protect the finish.  The end result was a somewhat dark gray that was a perfect contrast to the glossy black panel.  Again the photo doesn't do it justice.  I don't have a clue as to why it turned out dark gray.  I guess the matte clear enamel had something to do with it.  Don't know.

 

Note: The voltage adjust pot on the far right of the front panel.  This is a special pot with a screw driver slot and the shaft is too short to allow a knob to be attached.  While I wanted the voltage adjustment to be easy and on the front panel, I don't change it that often so wanted something that wouldn't lend itself to being accidently moved.  Perfect!   The photo shows a current of 20.1 amps at 13.8V.  My Yaesu FtDx-1200 was keyed up outputting 100 Watts when I snapped this photo.  At 20-35 amps continuous the power supply just loafs along.  It hardly breaks out in a sweat!  Groovy baby!

 

 

Ohhhhh look at those hefty 95 amp diodes!!!! Don't they just make you drool??  The heatsink is isolated  from the rest of the chassis and its H-O-T meaning there is 24V with a *LOT* of current.  If the heatsink gets too hot the thermal shutdown sensor will put the power supply into standby mode which means it will not output any voltage until the power supply is reset by the operator. Before my mods the designers put two  bridge rectifiers in parallel.  To compensate for this stupidity they put a thermal sensor on the heatsink to shutdown power to the bridges if they got too hot.  And why would they get too hot?  Because one bridge will end up pulling more current than the other, so this bridge will be the one likely to fail.  They were already using a center tapped transformer so there was no reason to not just use two high current diodes with their cathodes connected together and using the center tap for ground.  This is your basic full wave rectifier schema.

 

There are 8x 10,000 uF 50V electrolytics for filtering.  The original design had 6x 10,000 uF 25V electrolytics.  One day I was measuring some voltages and had my head not-so-far-away from the filter capacitor board (Not the homebrew one shown in these photos).  These old electrolytics had very low ESR's of about 100 ohms.  One of the old electrolytics puke while I'm measuring some voltages.  Blows up in my face.  The cap went off like a shot gun shell!  I kid you not.  The cotton and electrolyte just spewed everywhere at a very high velocity.  The photo below is the exact same capacitor.

 

 

Therefore please note the following and follow it:  Power supplies are dangerous! Electrolytic capacitors can blow up like little bombs sending hot liquids all over the place, on your face, in your eyes, all over your wife's favorite skirt and blouse!  The shrapnel from these capacitors can take an eye out.  The AC voltages aside (120VAC or 220VAC), there is sufficient current in a decent sized linear power supply that improperly designed and or built can result in a fire.  That can lead to loss of life or property.  Just because the voltages are low doesn't mean that one throws common sense and safety measures out the door on a whim.  This is seriously dangerous stuff.  A few years ago I was experimenting with a LM317 adjustable regulator.  This is a three terminal regulator that is capable of only an amp or an amp and a half.  I made a novice mistake in how I wired up the voltage divider for the voltage adjustment.  Somehow the thing at the end of its resistance made the formula as documented in the datasheet become undefined due to a divide by zero type fault.  Without warning, without excessive heating, without any smoke or any sign of trouble of any kind, the thing blows up. This was a older chip whose body was made out of ceramic.  The chip just explodes with a powerful POP and the ceramic body exploded into many pieces of high velocity shrapnel. It sounded like a .22 long rifle being shot.  Scared the living dog doo out of me.   I had glasses on, but some of the shrapnel hit me on the face next to my eye.  It cut the skin and was bleeding.   All that was left of the regulator was the three legs that was soldered onto the circuit board.  This was with low voltages and low currents. Did I mention this stuff is dangerous? 

 

 

The 2N5686 has a maximum power dissipation of 300 watts!!! Orginally there were six pass transistors which were 2N3772's (NPN, 20A, 60V, 150W) and I replaced them with four 2N5686's (50A, 80V, 300W).  The maximum power dissipation for the 2N3772 is 150W.   That's 1200W total (4x transistors x 300W) and so I was able to reduce the pass transistor count from six to just four.  These puppies are a little bit pricey but I found some at Newark Element 14 for about 7 bucks a pop!  Yeah, Yeah I can hear you say; There are more powerful ones; Yes there are.  Can I have your credit card so I can buy them?  Non?  C'est la vie!  Now it must be noted that the original transformer in the Pyramid Phase III PS-50 power supply will probably not be able to supply enough current for full 1200W output.  Each heatsink has a fan mounted on the outside rear that  is controlled by thermal temperature sensor mounted on the inside of each heatsink. If the heatsink gets too hot, the thermal sensor will turn on its corresponding cooling fan which remains on until the heatsink cools below its threshold 'on' temperature, which will then shut off. 

 

 

The original power supply had absolutely no inrush current control. When the unit was powered on the transformer would make a loud hum noise with the sudden application of current to it, so I added a MOV to keep this from happening.  Now it powers up completely quiet.  The original unit really had no AC spike protection on the primary of the transformer so I added a MOV (and a .01 1KV cap in parallel to the MOV) across the Netural and HOT and then a MOV from Neutral to GND and one from Hot to Gnd.  The original power supply had the 30 V regulator PCB power supply on a  PCB board but the PCB was just crap  so I had a nice barrier strip that I built the filters and the Regulator PS  as shown above.

 

 

There are two heatsinks with 2 pass transistors on each heatsink.  The original design was beyond retarded.  The pass transistors were configured as emitter followers in parallel with no resistors in the emitters  to equalize the current amongst themselves.  The designers ignorantly believed that adding series base resistors to each pass transistor would achieve equilibirum. Bzzzzt. Wrong! So I added a 0.05 Ohm 10W wire wound resistor in series with each emitter and dumped the base resistors was they were not needed.  These 'Ballast' resistors are what ensures that no one pass transistor will draw more current than the others. Like I said: The original design was beyond retarded.  It was utterly unreliable.

 

 

I was going to replace the original LM723 regulator PCB with a different LM723 design of my own. I even designed the PCB in Kicad and had planned on submitting the design to a PCB fab for some test boards so I would not have to endure the torture of building it on a pref board.  I then realized that I could repair the original regulator PCB (it had a bad chip) and then modify the design so that it would be more reliable.  The original design had two issues that were highly problematic rendering it unreliable:

 

  • The LM723 has a maximum Vc of 40V. The power supply for this chip is derived from a second low current secondary winding on the main transformer and after converting to DC and filtering this voltage is 30V unregulated.  Since varying loads can severely effect the voltage of this low current supply, transient voltages exceeding 40v are very possible and indeed are likely.  The regulator chip had no decoupling circuitry so I added a couple of low value resistors for decoupling  and a 37V  1W zener diode to keep the voltage stable and within acceptable tolerance limits of the chip.
  • The original design had the regulator chip driving the base of either a 2N3055 or 2N3771 power transistor. This transistor in turn would control  the output of the pass transistors. In this configuration the total drive current of the regulator would be less than the base current IB  needed to turn on the transistor it was driving.  This condition over time would lead to a blown LM723.  The solution is to replace the driver transistor with a Darlington pair transistor that has a resonablely high output current and a much lower IB requirement to control it.

 

The above two modifications are all that were necessary in order to make the regulator pcb more robust in operation so I could not justify the expense in time and effort in continuing with my own design.

Epilogue

Do not make the mistake of attempting to use a 3 terminal voltage regulator like a LM78xx, LM317, LM338 ETC. for your high current linear power supply.  The LM723 has been around for decades and there is a reason for this: It works. Sure the above mentioned regulator wanna be's can be made to behave nicely but good luck with that.  It wont perform as good as a linear power supply using the LM723 correctly!  I used to dislike this regulator but that was because I did not understand the chip.  I didn't understand the chip needs a regulated power source; I didn't understand the design needs to implement proper power decoupling techniques; I didn't understand the LM723 needs a good and proper frequency compensation capacitor.  I was a moron; I didn't understand that if  I read a datasheet of a component and didn't understand a thing I read, that there was no way I could expect a reliable cost effective design.  I  was deaf, dumb, and blinded to the fact that I blamed my lack of knowledge and understanding on the stupid chip.  I was a moron for having smeared, tarnished and bad mouthed a chip that has been around for over 40 years.  Oh the old timers who knew the real story though:  In their eyes I was a complete and utter moron.  I'm sure they would just shake their heads in utter horror upon hearing the foul and pathetic epitaphs  I bestowed upon this poor little regulator, blaming the poor little defenseless slab of silicon for my stupidity and outright ignorance.  Oh the horror! The horror!

 

The end result It's rock solid.  It will provide 30A continuous all day, all night and not even skip a beat.  Its regulation characteristics under a high current load are excellent.  The transformer is the hottest component but there is more than adequate air flow across it to keep its maximum temperature well within an acceptable range of operation.

In other words:  This power supply kicks butt. What would I do different?  Replace the transformer with a more efficient toroid transformer.

 

 

Upcoming Projects:

  • Running QRP from a pair of Eimac 3-500Z's.
  • Modulating the filaments of a  pair of 3-500Z's for unique cold cathode RF emissions.
  • Filaments burned out on your expensive HF Amplifier final PA tube?  Tutorial on how to use a hair dryer to blow new life back into your no filament cold cathode PA tubes. Heat the cathode with your hair dryer.
  • Get the most bang for your buck: Powering a crystal radio via 3 phase 440VAC.
  • Spark Gap controlled crystal VXO.
  • The Bounty BiPole antenna: A radio frequency quicker picker-upper.
  • The ham radio Clapper:  Clap your hands to automatically QSY and call QRZ.

 

Other things on the Burner:

  • Ham Radio bargins at Walmart.
  • Assisting my dog in passing his Extra class exam.
  • Classical baking using your HF power amplifer  as an oven!  Free recipes.  Be sure to try the resonating pancake recipe!
  • Using your HF power amplifier as a moonshine corn liquor still!
  • Using long waves to stimulate marijuana growth.  (Open beta: FCC license required).
  • Pine tree vertical.  Oak tree tribander.  Hickory nut log periodic array.
  • How to use hickory nuts, peanuts and pecans as a HF  through UHF 50 ohm dummy load.
  • Replacing all the electrolytic caps in your vintage HF power amplifier with Lyden jars.
  • Radio controlled pigeons in the 50Mhz band.
  • Training squirrels to help with placing your antenna wire in trees and other tall objects.

 

Back to Top

7710481 Last modified: 2016-11-21 13:39:08, 56763 bytes

Login Required

Login is required for additional detail.


Apply for a new Vanity callsign...

You must be logged in to file a report on this page

Please login now...

QRZ Logbook Summary for - KG5NII
Latest Contacts for KG5NII at QRZ.com
dedateband mode grid Country op
S01WS 2016-12-07 17m CW IL46rd Western Sahara Sahrawi Amateur Radio Club
T77C 2016-12-07 17m CW JN63fw San Marino ANTONIO TONY CECCOLI
JI3DST 2016-12-06 17m CW PM86TF18 Japan TAKESHI FUNAKI
JA1GQL 2016-12-06 17m CW Japan
J6/KI8R 2016-12-06 17m CW St Lucia
JH1OCC 2016-12-05 17m CW QM06ha Japan HOSONO YUUICHI
F8BBL 2016-12-05 17m CW IN94su France Laurent DUMAS Lor in cw
OF9X 2016-12-05 30m CW KP17uw Finland Santa Claus World
JR2UJT 2016-12-04 17m CW PM85rm Japan Kenzo Suzuki
HP1RIS 2016-12-04 17m CW FJ09GA24 Panama Ricardo Schwarz
HK1X 2016-12-04 17m CW FK20fj Colombia Pedro Claver Orozco
TF3Y 2016-12-04 17m CW HP94bd Iceland Yngvi Hardarson
JO1WXO 2016-12-03 17m CW Japan Makoto "Maco" Shibata - M0WXO
JF1IRW 2016-12-03 17m CW QM05 Japan TAKEHIKO URIU
JA7QVI 2016-12-03 17m CW QM08ka Japan TAC TAKASHI HIRAMA

Book Totals: 151 qso's   17 confirmed Get a free logbook at QRZ.COM

ad: giga-db
Copyright © 2017 by QRZ.COM
Mon Mar 27 06:45:11 2017 UTC
CPU: 0.129 sec 127414 bytes mp