Home brewing a ham station

The last half of July and August so far has been hotter than the henge's on the gates of hell here in Central Tennessee, high nineties to low hundreds most days an not much relief at night. My shop is a 12' by 20' prefab metal building and the insulation is inadequate to deal with such temperatures. I do have a small window air unit and with reasonable outside temperatures it manages but not for the last month to six weeks. Needless to say I have chosen to stay out of the shop on hot days.

That said , I have had some problems with the digital frequency readout, initially the problem turned out to be a damaged TCXO. I have no idea whether it was a bad unit to start or if I accidentally damaged it, but it failed during the calibration process, no joy! However , Neil at AADE helped with the problem and this week I was able to complete the assembly and set the If Frequency into memory. As I stated in an earlier post the receiver now works very well on 80 meters. although I am less than thrilled with the IF and Xtal filter. When the  digital frequency readout has no input from a converter crystal Oscillator the display reads from just below 3.5 MHz to just over 4MHz. When I adjust the BFO crystal adjusting capacitor. The frequency is displayed followed by either LSB, CW , or USB depending upon where the BFO is set in relationship to the IF frequency. I am still not sure how the converters will  display on the readout , got to figure that out before I run out ad buy a bunch of rocks.Well it is two days later and we finally got a break in the high temps,
Made considerable progress on the rig. Got the counter mounted and a face plate fabricated and mounted as well. Still have some tinkering to do getting the counter to read the correct frequency when the converters are added, Here is a pic of the rig as it is now. The band switch when added will be located below the S meter and above the on /off pull switch with no knob. I hope to score a suitable Knob at the Huntsville Hamfest this coming weekend.

Tags: Untagged

Did some fine tuning on the receiver today. I was pleased that the receiver knocks your ears off when tuned across my signal generator using its lowest setting, (250 milli-volts to 50 ohms) using about 2 feet of  wire lying on the bench as an antenna. All symptoms were good , there is a noticeable increase in the noise level when an antenna is connected.there is plenty of IF and audio amplification everything should be hunky dory right? Well not so, I waited until this evening when 80 meters came alive again and tuned across the band, I heard a few signals but not as many as should be. I decided in the interest of science (and air conditioning) to bring the rig into the house and work on it from my operating bench. Big difference, the band was filled with signals both sideband and CW when connected to my operating antenna (wire in the trees). I obviously have a problem with the shop antenna. I wonder how many of the module rebuilds were unnecessary, oh well, my learning curve may be low but it does curl up eventually.

I tried another version of the crystal filter, this time a three pole filter designed for sideband use. I am not pleased with the bandpass characteristics but it may be about as good as I am going to get with these crystals.  Lately I have been leaning toward trying the sideband filters I mentioned earlier 5.6 MHz That would only require a VFO change of some 300 kHz and that is easy enough to do. Perhaps I will give that a try but first I want to rework the four pole filter and see if I can work out some of the rough spots in it I am of two minds about incorporating an rf amp in the receiver. Normally judging from tonight's listening session I don't need one but there are times when one could be handy.. Perhaps one that can be switched in or out would be nice. 

Tomorrow is another day but unfortunately there are appointments. One of the things about getting older other than the obvious is the need to watch after ones health more so than when we were young, of course if we had looked after it better when we were young then perhaps the need would not be as great  who really knows... the thing to do is LIVE forever or Die trying I guess. 

an added note, I just went back and ran a spelling check on all the entries of this blog, For those who may have struggled through them let me offer my sincere apology, I do know better...I am sorry for putting you through all that misspelling.

Tags: Untagged

It has been almost a month since my last entry in this category. No excuses except to say that Life has a way of getting in the way and there has not been any news worthy break-through to write about. A lot has happened, Progress has been made and a lot of blind allies have been explored.  At this point I think I have rebuilt every module in the rig at least three times except for the VFO. some builds worked badly , some not at all and a few worked well. Some of the mediocre ones after trouble shooting have turned into keepers and ones I thought to be keepers early on have turned into "also ran". If there is anything that stands out in attempting an effort like this is how much there remains to be learned, even about circuits I though I understood well. Some of the circuits that I have rebuilt were brought about by me having a better understanding as to how they should work and usually the rebuilds have been worthwhile. Here is where we are now

I now have a functioning 80 meter receiver. It has much room for improvement but it is a long way from where we were at my last blog entry. The copious bird calls I reported have been tamed, there are no more shirks and whistles and ear splitting heterodynes heard in the phones. Those were eliminated by careful bypassing, shielding and  placement of modules.  The bandpass filter that I settled on is a version of the bandpass filter used in W7ZOI and K5IRK's Progressive Communications Receiver (QST Nov 1981)  The secret to building this filter, if there is one, is to read the instructions carefully and measure every component used in the circuit. I used combinations of standard values of capacitors to arrive at the values specified. During the process I also discovered I had been misreading the number of turns called for, on the coils in the tunable section of the filter. When the correct values are used it is amazing how well this filter performs. As an experiment I built a version of this filter using a circuit board (commercial) and a bag of parts sold at one time as a kit for this filter. The kit used standard closest value 5% capacitors and not the values called for in the original article for example 4700 pf caps were used for the two 5000 pf caps called for in the article. While it is possible for 5% caps of this value to be very close to the required value these were not and the end result was  a very poor filter with a lot of insertion loss and weak peaking. I built my filter Manhattan style and used capacitor combinations (as measured on my ALDE LCII Meter) equal to the values called for. Likewise the coils were wound and measured for correct value. The result is a very good front end pass band filter with under three db of insertion loss ( if my measurements are correct)and with the peaking capability off setting the losses. What was especially satisfying is that I now know why previous versions of this filter that I had built, had sometimes performed so poorly. 

I had problems deciding upon an  IF amp circuit, I built several all of which worked after a fashion but most had issues.  Two were built around MPF131 dual gate MOSFETS and two around cascode arrangements. The MPF131 units were OK , however one had far too much gain and both had stability issues. The red LEDs in the source biasing were fun to look at. The stability issues could likely been overcome with use of Ferrite beads but I had none on hand.  Same story with the cascode circuits using J310 FETs and 2n3904s in cascode. TOO much gain in both cases and not enough control. I Finally settled on the cascode circuit used by Frank Harris (From crystal sets to sideband ) Available as a free download on line ) using 2n3904s only. It is easy to build, works well, is reasonably quiet,is easily controlled and seems to be immune to oscillations. Until I find something a lot better, it is my choice. 

I am still not satisfied with the crystal filter. I measured and tested 50 crystals (5.905 MHz a ham fest lot) for several parameters using several pieces of home brew gear (that I had to build in order to use) and found several sets that appear to be suitable for filter construction. I chose the Minloss arrangement (Cohn) using matched rocks but it seems that they either do not have adequate q for filters or my home built test gear is not suitably accurate to measure them correctly as the filter seems to have excessive loss and is a bit more narrow than I expected. I am using a four pole filter designed for about 2.4 kHz bandwidth. I may redo it with more careful attention to layout and see what happens.  More on that later. If all else fails I have acquired a couple of Drake upper and lower sideband filters in the five mhz range that I can use although I wanted to use a home brew filter in this rig.

That is where we are now, I will try to get some more photos when I have a chance but for now I am concentrating on building the best 80 meter receiver that I can build after all the band converters will only be as good as the rig they convert to.      

Tags: Untagged

Got around to actually hooking the modules together and doing a road test.It works in a sense, sounds something like an aviary at the zoo at feeding time, more squeaks, squawks and whistles than a Mockingbird on drugs. One of the problems is that the IF system has far to much gain.  That is fixable several ways but after all is said and tried I may just go for a system with at least one fewer stages. At this point I am not happy with the front end filter, it introduces to much loss  and I am not at all certain about the Frequency scheme I have chosen,  The VFO works fine as does the PD and audio amp and the mixer is satisfactory but running with the barn door wide open ( no filter installed) there are too many birdies for my liking. I am hoping that an IF filter will do away with some of the more annoying chirps and whistles which should be beat notes of harmonics well outside the desired pass band pf the receiver. Time to build a filter before rebuilding the various modules, more later!

Tags: Untagged

Checked out the modules so far and they all seem to work except for the IF system. By connecting the output of the first mixer to the input of the Product Detector  I can detect signals from my signal generator  Of course there is far to much loss in the diode rings to enable on the air signals to be detected although both  PD an First Mixer work perfectly well as DC receivers when connected individually to an antenna and  to the audio amp. The VFO and BFO are operational as well. Still I need a lot more gain to overcome the mixer losses and bring the weak on air signals up to a usable level. To that end I started on another IF strip. This one was published back in the 80's by a french ham F6CER and published in QST Magazine (August 1981) under De Maw's by line) It uses three dual gate MosFETs as the gain element and an AGC system using MC1350P and 741 ICs plus a couple of FETs. Here is a pic of my effort using Manhattan style  Construction. As I build, I like to use a red ink pen to highlight the parts on the schematic as I go. I find this helps to eliminate omissions of parts and allows me to start and stop as the need arises. The transformers use here are units 10.7 MHz units salvaged from defunct transistor FM radios. Here is a closeup of the partially completed if strip this is the section with Q3 and T3 and T4. No attempt has been made to miniaturize this section since doing so would make testing and repairs difficult and I have no need to do so in this receiver. In the original article 80673 or 3N211 MosFETs were used, at over $6 each when you can find them I elected to use MPS131s at $1.75 from Dan's Small Parts and kits. Tags: Untagged

Pat , understanding young lady that she is, presented me with a pair of BOSE QC15 noise canceling headphones, Joy to the world! They are comfortable to wear and the noise canceling circuitry really works. It is raining like crazy here today and the noise level must be 20 Db over S9 here in my home office, Hard rain on a metal roof if you have ever heard it you know what I mean, with the phones on I can barely hear the rain. They are not cheap but they are very nice indeed.

Worked last evening on mounting some of the circuits to the chassis. The band pass filter (80 Meters) was usable but had twin peaks a couple of KHz apart. As I said it was usable but annoying. The literature suggests that this is  caused by over coupling between stages. Since I am using toroid coils ( self shielding)  mounted perpendicular to each other I did not believe the over coupling was via the coils, A small value NPO trimmer as a coupling between stages had allowed me to bring the peaks much closer together but still visible on my scope.  The trimmer measure 5.7 PF minimum capacitance.  It was replaced it with s small 2.2 pf tubular cap (ceramic I think) and that solved the problem. Now a single peak and a fairly sharp filter that tunes from just below 3.5 to a bit over 4 Megahertz. It is not broadband tuning but I don't mind that , I prefer the additional selectivity up front.

 Played around a bit with the BFO circuitry again as well,  Using the switched BFOs for upper and lower sideband was not a problem but neither is suitable for CW and making them variable or adding a third BFO crystal was an over complication that doesn't well with my idea  of a simple serviceable receiver. I found that by using a circuit that has inductance between the crystal and the tuning cap one can achieve three or four kilohertz of frequency swing. A small trimmer in parallel with the main tuning cap sets the lower swing edge. There is some interaction between the capacitors, the inductance and the crystal. One also must keep the inductance low enough that it doesn't dominate the circuit. I've arrived at what I believe will be satisfactory and the BFO will be tunable from the front panel for both sidebands and CW.  The shielding will be added after I determine whether or not I need more output power from this stage. I am not sure this will drive a diode ring PD with enough oomph!

Tags: Untagged

Got  some quality time in the shop today,  Completed the audio amp and Product Detector. AS I suspected the 5 Db conversion loss shows up in the audio drive and to counter that I added a small audio preamp using a single 2N3904. which more than makes up for the loss and gives me a few extra Db of drive.  In the photo, the product Detectior is on the left and the audio amp is on the right, the space between the two is where the audio filter switching leads and volume control connections will go. There is plenty of extra spce on the board and I may decide to construct a pair of audio filters on the board , perhaps one for around seven or eight hundred herts for CW and a wider one for sideband.  The BFO circuits and housing were completed as well, I said circuits because I elected to use two xtal oscillators which can be switched for upper or lower side band. They have yet to be modified for CW off set  whuich I will probably do with either a diode switch to the capacitors or perhaps I will just add an additional oscillator dor that purpose.  Here you can see the two circuits , side by side, they are switched by application or removal of operating voltage through the two feedthrough capacitors on either side of he audio socket. The cover is held in place via the two screws lying in the clam shell lid. The BFO will be mounted below the chassis while the IF amp is to be located above the chassis . I do not want BFO energy finding its way into the IF chain. Speaking of which, I located a schematic and article in the QST archives (the archives alone are worth the price of membership) which depicts a dual gate FET if Amp and AGC system. While I am not giving up on the cascode amp, I am considering build the FET amp, I will let you know what I decide...Enough raving and ranting for now. Tags: Untagged

Its has been several days since my last post , however, now that the various "meeting" gods and the "Lets all get medicated" gods have been appeased for a month or so I can now get back to business. I am currently working on the Product Detector/ Audio Amp section. As the product detector I am using an ADE-1 Surface mount Diode Ring. The audio for now anyhow is a generic LM386 following a 2N3904 audio preamp. Both these are pretty much run of the mill and I won't go into a lot of detail. After all the purpose of this blog is to just let others know what I am working on. It is not intended as a "How To" blog. Once I get the PD/Audio section finished and tested, I can do some meaningful testing of IF amps and BFO etc.   AN interesting aside, I recently was listening to the local state evening net where everyone more or less says I am here by checking in by counties. Signals were pretty fair but the usual suspects were all running their full gallon rigs and some of them splattering all over the place. I decided to check in using my Yaesu Ft817 on battery power and using an unmatched antenna. Output power about two watts, they all heard me and the net control gave a 5 by 7 report. OK, so I wasn't 40 over 9 at the other end but with 2 watts who cares. Tags: Untagged

 April 17 2011:  Yesterday and today, I bit the bullet and built an IF/AGC Circuit. I chose one modeled on the cascode If amplifier in EMRFD but with some personal tinkering that the original circuit did not have. Each of the JFET sections was tested as it was completed. A signal from my Home brew signal generator with a step attenuator in line (also home brew and based on QST article of the past.) was applied and the output was observed on a scope.The circuit was powered from the receiver power supply and the 5 volt output was used to vary the gain control. Each section worked after a bit of tweaking as expected. I did not have the coil cores cores used in the original article which was designed for 9 megahertz, so the coils were scaled for 5.905 megahertz and  FT37-43 cores were substituted instead. They seem to work satisfactorily. Some of the things I did differently were as follows, a single string of three diodes were not used to bias the JFET sections as per the article. Instead, the first two sections are biased by a two diode string in series with an optoisolator. The plan is to see if I can drive an s-meter  from the transistor output of the Optoisolator. The chip I am using is a four pin chip harvested  from a battery back up system board. The third JFET amp is biased by two small LEDs and an 1n914 diode. Mainly This was done to observe the changes in current without having to hook up a current meter each time a test is run. It is interesting to see the LEDs increase and decrease in brillance as the AGC voltage is varried. ( Doesn't take much to entertain me) With the JFET sections working correctly The differential pair was constructed using 2n3906 tyrannies. With it in place and the output connected to a fifty ohm probe to my scope. There appears to be plenty of gain when a low level signal is applied to the strip input. I calculated gain control variation in excess of ninety DB. The QST article states that AGC variation of up to 100 DB is possible with this circuit and that prediction appears to be on the money. Finally the AGC amplifier was constructed but awaits testing for now. Manhattan style construction was used throughout the circuit.  Update: It appears I have a distortion problem with the differential pair amp. Input is a pure sine, output looks like hell ... got to do some more trouble shooting on this one.

 

Worked on several aspects of the receiver today,  Since I prefer to use the power supply in the receiver that it is tested on I built a Power supply with 12 and 5 volts output both regulated. I used a Rat Shack transformer The one rated at 25 volts center tapped and capable of a couple of amps of current, this should be more than adequate for this receiver. The transformer was purchased new , everything else came from my junkbox. The digital display was test fitted to the front panel and a gasket was fabricated to ensure the display does not short  to the panel. Also checked that the  counter assembly would clear the VFO housing.  I chose to use a big display as it is easier on the eyes and at this point in life, easy reading is important.Here is a frontal view of the display and meter in place.The meter is the Drake meter with the face redone. I used the original housing to hold it in place. I will have to use a sheet of aluminum to cover the unused holes on the Front panel for appearence more than anything else.

 

 

Except for the transformer all the power supply components are placed under the chassis. The large filter capacitor is a salvaged unit of the Snap-in persuasion that I think I harvested from a dead Surge protecter battery backup unit used on computers. I had to solder leads on the snap-in prongs and a drop of super glue holds it in place. The regulator just below the capacitor is a twelve volt unit and the bottom one is 5 volts. 

 

 

 

 

 

 

Worked on 80 meter input filter as well. After several attempts to build and use a commercial filter board for the W7ZOI progressive receiver, I gave up the attempt. Although the high pass portion of the filter worked just fine the tunable section had very high loss which I was never able to get below 20 db or so. I don't understand why it didn't work, all of the parts were measured with an AADE LC meter which has been checked for accuracy and all the coils and caps were withing 5% tolerance of the values published in the original article. I have read where others have had simular problems using the commercial circuit boards. Perhaps I will experiment with the board later but for now I built a double tuned input circuit which aligns reasonably well ( no double resonance points) If Broadcast band interference becomes a problem I will put a high pass filter ahead of the tuned section. As  side note., I received two AD8307 chips I had ordered on line. The two 8 pin Dip chips arrived in a 6 X 4x12 inch box. Makes one understand that many of he parts sellers out there are really not set up for small orders.They could easily have fitted an order 500 times larger thn mine in the same box.

Tags: Untagged

Seems to me that one of the most vexing problems associated with being an old ham is boredom. Most of us who have been hams for  very long have worn the shine off in more ways that one.  I have gear that I could only dream of owning when I was a beginning ham. Problem is most of it is about as exciting as vanilla flavored rice. Most of the nets of which I am aware are pretty much the same. One thing I have always enjoyed doing is building ham gear. Not because it is better or cheaper or even better looking because in my case it is not. I build about as many circuits that fail as work. But when they do work, the thrill is back again. I find that as I learn more , trouble shooting becomes easier and more and more circuits seem to eventually work.

What I don't need is another ho-hum radio with a zillion features that I will never use, and  more computer power that NASA had when we went to the moon. I have a perfectly good Kenwood that is showing its age but still works perfectly well, I have a new or nearly so, Yaesu that one needs a PHD in electrical and digital engineering to set up. it works well too.  Problem is both are about as exciting as an wrong number ring up on my cell phone. I need to recapture the thrill of ham radio. To that end I shall build a complete station. It won't be easy nor quick but I shall do it. Perhaps one day I can pass the Commercial rigs on to new hams who still can see their shine.

I recently attended a small ham fest in my area and in addition to about 50 lbs worth of power transformers I scored several junker Drake TX4 C and B transmitters along with one badly damaged R4c receiver , I may repair it and one of the transmitters eventually but that is another blog. I chose one of the junkers as the platform for my receiver. I harvested a wealth of parts from that rig and any of you Drake purists out there don't get your panties in a wad,  the junker was long past  restorable. I did not destroy a valuable and irreplaceable artifact. The chassis required some work , the frame on the Right side required straightening, To cover the holes where the tubes and other components had been, I cut a piece of aluminum flashing to size and glued it in place using 3 M spray adhesive. It looks good from the top and no one will see the underside anyhow so big deal, Right? This should serve as well as a brand new store bought chassis at very low cost indeed.  My design criteria was that is that I will build a receiver which covers the 80 meter band and use crystal controlled converters ahead of it. Since I plan to use some 5.905 MHz crystals  that I have on hand to build the Crystal filter, a VFO that tunes 9.4 to 9.9 Megahertz. is required. Stability and band spread can be a problem at higher frequencies but I seem to have gotten around those issues by using an very stable NE602 as the oscillator placing it inside a decent shield and using series capacitors to achieve the band spread I wanted (500 KHZ). Regulated voltage is a must and I used it. To tune I use a salvaged Broadcast band capacitor with a built in reduction gear, I couple this with a 7:1 Jackson Brothers reduction drive  for a nice tuning rate. The tuning is of course more compressed at one end but since I am using digital readout the problem is reduced.

 

Tags: Untagged