Sunday, December 31, 2017

2018 ~ The Year of SSB Transceivers

Clean Off the Work Bench ~ New 2018 Transceiver Projects!

1/18/2018 ~ Now you found it -- How to Fix It?

I was so excited yesterday that I found the source of the problem --it was not the rig per se or a test of my schematic reading or soldering skills --it was RF in my back yard!

One comment on the you tube video --how do you fix it as this will also have implications on the IC Transceiver build. My XYL suggested moving to an upscale neighborhood that has antenna restrictions. Should note for over 50 years she has absolutely hated the hobby so her response is quite clear -- make me QRT.
 
In the Page 104, SSDRA design had as a front end a simple coil and a cap for  20 Meters. Whereas the 80 Meter filter uses a Cohn type tuned network. So one possible fix for me since I don't operate CW or Digital modes is to come up with a Band Pass Filter in place of the simple tuned circuit that would be centered on the phone band and with care could offer a 15 to 20 dB reduction to signals below 14.150 MHz. That is possible.
 
Another would be to design a trap "notch filter" centered on 14.075 MHz which of course if I did operate digital would kind of makes things hard for any digital operations; but not incompatible with my operating mode desires.
 
This is a first for me but it does appear to be limited to 20 Meters.
 
73's
Pete N6QW

1/17/2018 ~ Trouble Shooting Mystery Solved

So Ok, yesterday I determined that some evil external force was de-sensing the front end of my new (rebuilt) transceiver. But the wild goose chase was invaluable as I had a real chance to dig into what I had done some 35 years ago where the documentation of what had been done got lost in one of the 10 moves I made in that 35 years. I even rebuilt the active mixer --so all is not lost. My detailed recent journey did affirm I knew what end of the soldering iron was hot --even 35 years ago. BTW I did have to go before the FCC Examiner to get my Extra --no box tops here!
 
This morning I had an inspiration and that was to connect my Rigol scope to the rig's antenna with the rig turned off and just see what I could see. Initially there was a bit of background noise and then the screen filled with signal. I had the counter function engaged and quickly noted that the frequency of the strong signal was 14.07475 MHz and it was really strong. I observed several cycles to affirm and verify what I first initially thought this morning. Boom someone close by is operating in the digital mode --and I mean close by like maybe 800 feet. It was a ham on the next block and just recently I noted he sprouted a new antenna on his roof.
 
Next I turned on the rig and tuned down to 14.07475 and still with the scope connected across the antenna spotted the screen filling up and at the same time the rig's S Meter showed about a 30 DB /S9 signal. I watched that for several cycles and indeed that was the culprit. Keep in mind that was 30 DB/S9 in a de-sense mode. My ham friend must have gotten a new radio toy for Christmas as I have never had this problem before.
 
For your amusement and amazement I made a video of this morning's solving of the mystery. See below. There was a method to my sharing this adventure. In troubleshooting a rig, often you have the data right in front of you -- the problem is one of how to connect the dots. It took me two days but was an invaluable learning experience for me -- there was nothing wrong with the rig! The problem was the rig was being subjected to very strong signals in close proximity -- connected to a beam only made the signal stronger!

 
 There we have it -- another ham is the problem. Feel a lot better now.

73's
Pete N6QW

1/16/2018  Tribal Knowledge Trouble Shooting (1/17 Update)

As we prepare to build the IC Transceiver I wanted to spend some time on a real world rig
problem that directly applies to the IC Transceiver project. What to do when something is not right? In the most recent transceiver project (in 2018) I rebuilt something that I had originally built in about 1985.
 
The 2018 Reborn 1980's rig. The LED next to the Meter Glows Green on Transmit. How Cool?

There is a story behind this rig and it all started with the Solid State Design for the Radio Amateur (SSDRA). You have often heard me say that I prefer this publication over EMRFD. Rebuilding this project affirmed that position. If you go to Chapter 5 page 104, you will see a two band Superhet (20/80 Meters) which uses a 9 MHz IF. When I looked at that schematic, back some 35 years ago my first thoughts were why can't this be made into a transceiver --which it was.

(For those who do own a SSDRA there is a close parallel to the page 104 design and the K1BQT 75M Transceiver in that both use an active mixer, both use the MC1350 and both use the same AGC circuit. Don't overlook the diode steering in K1BQT's rig to steer the signals into/out of the filter. Undoubtedly there has to be parallel influences at work here! So spending some time with SSDRA is most useful before building the IC Transceiver. Like I said my copy of EMRFD is a great bookend!)
The Front End active Mixer is a 40673. Having some 3N211's I used that for the active mixer. The plan was to use diode steering so that the Rx signal would be diode steered into the IF stage. The other half of the steering would be the output of a diode ring balanced modulator with a 741 Op Amp microphone amplifier.

Thus signals passing through the IF would either be the Received signals or the low level Transmit signals. Based on a concern I had at that time with using the KVG 5 Pole filter I added a simple 2N3904 buffer stage ahead and following the filter. The idea was both in SSDRA and the seminal ON5FE 1970's IF Module. From there just as on page 104 the signals were routed to the MC1350 IF amp and thence on to the singly balanced diode ring. Noteworthy I kept the AGC circuit with thoughts of using it with ALC like in the IC Transceiver. [That may not actually be a good idea --will explain in a later post --but the ALC control level is affected by the level set for the AGC.]
So now looking at the singly balanced mixer (the OBTE are left to figure that out) I added more diode steering to the what is normally the LO Port. Thus I had diode steering of the BFO on receive and more diode steering of the output of the single balanced mixer to the audio amp.
But on transmit the LO port was now fed the LO signal from the VFO and the output now steered to a band pass filter and then on to a 2N5179 RF stage. [Hayward loved 2N5179's]. From the 2N5179 the RF signal was passed to a driver stage and then on to an IRF510.
So the changes I made in the upgrade are I ditched the diode balanced modulator and removed the carrier oscillator circuitry. In its place I installed an ADE-1 and the BFO signal is now supplied by a CLK2 from the Si5351. The LO signal (above the incoming at 23 MHz) is supplied by CLK0 of the Si5351. I was unhappy with having a single balanced mixer on today's bands so I removed all of that circuitry and installed a TUF-1. [As an experiment with some surplus DBM's there is now a TFM-2P installed there which is a TUF-1 on Steroids.] The RF Driver is a 2N2222 driving a BD139 and the final a IRF510. Needless to say NO Analog VFO's or Crystal BFO as these technologies are out of step with the rigs of today!
Active Mixer now a 3N209

TFM-2P DBM, MC1350 to the right of the DBM

BPF and 2N5179

2N2222 and BD139 Driver Stage

IRF510 Final Amp bolted to the back wall of the case. Having a CNC sure is nice!
 

 

So all was going nicely with the new rig when suddenly I would encounter an intermittent condition on receive where the signal level (more like volume) would be reduced. The Transmit signal appeared to be not affected. So it was a receive problem.
My first "noodling effort" was to focus on the active mixer stage. Why? --well everything after the mixer is more or less common to transmit and receive. The only item not so other than the receive mixer would be the audio amplifier chain.  But I concluded, that the only  possibility since it worked well on transmit, was the active mixer.
 
The trimmer caps originally installed were orange in color and had a range of 15 to 60PF which I found in other installations had a finite life of so many rotations. I thought perhaps they were failing and would change value thus "untuning" the tuned circuits. So I decided to replace all of the components in this stage. I replaced the 3N211's with a 3N209 which was actually sent to me by the man himself, W7ZOI. With everything replaced I fired up the rig and all worked. The only thing this proved is that 1) I did not make any wiring errors and that 2) I do know how to read a schematic.
 
My euphoria was short lived for the problem returned. Again transmit was not affected only receive. I guess I should have read the tea leaves more closely as when the condition occurs the signal drops but does not go away.  It just loses signal strength. I prodded and poked components all along the receive path yet the condition did not abate.
 
A little more about what I am seeing (hearing). It is a cyclic thing where the signal weakens for a short duration and then gain comes back full on. Hmmm could there be a bad component in the AGC circuit that was cycling the MC1350 so that gain was being reduced intermittently. I was mystified. Again transmit was not affected. Disconnecting the AGC had no affect on resolving the cyclic de-sense.
 
So some other things I tried -- looking at the whole system. What about an antenna issue where maybe there was an issue with coax/beam (20 Meters)? So then I switched antennas to the dipole which is actually 3/2 wavelengths on 20 Meters. Boom same problem so that led me (falsely) back to the rig. Should have thought a bit more about that conclusion --if it loaded OK on transmit with no issues then it was not the antenna. I was bummed out as I just could not see what was the problem.
 
Well I decided the best course was to lay aside the rig for a bit and give this more thought. I heard a few stations on and so I put a different 20 Meter rig on the air -- Boom and double Boom --Same problem! Wait a minute. Tried a third 20 Meter radio -- same problem and so it was with a 4th 20 Meter Rig. Additionally one rig is a 20/40 Meter version -- there did not seem to be a problem on 40M. Thus the de-sense signal is not totally broad band. Thoughts floated back to when as a kid I built a TV killer which blanked out one of the TV channels so I didn't have to watch the Ed Sullivan show but avoided impacting the other channel which was the Colgate Comedy hour. Hmm was some one in the neighborhood trying to stop me from operating on 20 Meters????
 
So now the pieces were beginning to fall into place. There is something new in the neighborhood that is cycling on off and is a very strong signal that is causing my rigs to "de-sense".
 
So while replacing the active mixer may have not been necessary it does make me think it was not all bad. Tomorrow I will do more snooping around the neighborhood, On my morning walk I will bring along the rig with a battery pack and a short antenna to see if I can pinpoint the culprit.
 
All of our utilities are underground and it was not my neighbor welding in his garage as I know when he is doing that. Several years ago we had the "pot lady" growing weed in her garage and it was the cycling of the grow lights that made RF noise. We didn't get any new appliances so it is not in my QTH. It has to be a pretty strong field to cause a de-sense of the front end. I am having a hard time internalizing that the signal would be so strong as to essentially overload the rigs. Hearing RF interference (pot grow lights) is an order of magnitude different from overload.
 
I want to also re-verify that all the rigs are affected and that I have not missed anything. Oh I have three power supplies -- the same problem with any of the supplies. So I eliminated the power supplies as the root cause. Time now to rest and reflect.
 
This is a classic "Easter Egg Hunt" looking where all the eggs are hidden. I have never seen anything like this before.
 
Will keep you posted. But this could be a problem that might have been encountered with the IC Transceiver and I would have been chasing a bad circuit/component/wiring error when it was something entirely different. Sometimes the obvious is not so obvious --why did I immediately think there was a problem with the rig? Good question to always ask first!!
 
73's
Pete N6QW

1/13/2018 ~ More Considerations: To QRP or Not QRP That is the Question?

For those who may not know this I am a member of the QRPARCI Hall of Fame. I believe my selection was quite accidental and most likely those who are board members and regular members probably are wondering why as well. I am not a closet QRO guy hiding out as a dye in the wool QRPO guy; but freely admit my penchant for something more than 5 Watts!
 
There are many choices in our hobby so don't get snooty about having a DXCC certificate having done so running no more than 5 watts. All I can say is congratulations; but also be a ham and recognize others may not share your interest.
 
My interest in QRP is not driven by a desire to put a rig in a back pack, don foul weather gear and climb to a local mountain so I can operate in a driving snow storm and claim that I worked a station 200 miles away on 200 milli-watts SSB all the while braving off the elements.
 
Let me tell you why so many of my rigs end up being QRP. Early on I tried to build rigs that put out 50 watts from just the rig itself. For the most part I was terribly unsuccessful! Some how everything worked FB, until I got to the final amp stages where regularly there seemed to be problems with RF Feedback, bias issues and the heat problem creeping over to other parts of the rig. Was I being punished for running more than 5 watts?
 
What I did find is that almost 100% of the time everything worked great up to the 2 to 5 watt level coming out of the driver stages. Thus a solution -- build rigs that reliably put out 2 to 5 watts with no hiccups or burps and then add an outboard amp. So QRP level rigs for me was not to be a QRP enthusiast; but rather as a solution to some technical problems. I frequently run 600 watts with my homebrew rigs with no problems. On rare occasions I run my rigs straight through to an antenna running just 5 watts just to say I can run QRP but that is not my usual approach.
 
Now to the IC-Transceiver. While I intend to follow my usual practice of having an output in the 2 to 5 Watt range just to say it works flawlessly, there are some amps that look like they can be built in the rig itself without problems. One of those is a 20 Watt amplifier from K5BCQ which can be found at this link (scroll down as there are a lot of other projects on the page). Keyes and John know their potatoes so it is an amp that is "bullet proof".
 
 
So this is another piece to IC Transceiver. At the 20 Watt level you will get an order of magnitude (that is 10X for the OBTE) of greater contacts versus 5 Watts and thus a worth while approach. The 20 watts will also drive many of the MOSFET Solid State Amps to the Kilowatt level. But that is a project for another time.


Stay Tuned.

73's
Pete N6QW

1/12/2018 ~ Design Considerations & Decisions

For the most part we will be building the basic K1BQT transceiver or I should say a major part of the design --very likely 51% but not all of it. Today's post will cover the original design and what we will be changing and how we will change it. Firstly the original article was for a single band (75M) and employed a rather exotic way of getting 30 watts (a 28 VDC RF MOSFET) to the antenna. Some 33 years later we have at out disposal many tools and techniques that weren't even on the drawing boards in 1985. We hope to do this in an ordered fashion:
 
  1. The original design used an analog VFO and crystal BFO. The obvious 2018 change would be a Si5351 which will handle both functions and costs $8 as a complete board. The control of the Si5351 will be an Arduino Nano and the display a color TFT. With a bit of shopping this can be had for a $20 bill. The Arduino using built in capability enables ready band changing and control of other functions --so a bonus. More on this to come.
  2. The transmit mixer stage which is a MC1496 has a broad band output which is fed into a gain controlled 40673 Dual Gate MOSFET (DGM) . If you look at the schematic you will see a trim pot connected to Gate #2 which is labeled "Drive". That is a nice feature --we will keep it, But we will first start by ditching the 40673 (unless you have one in a bin) and replace it with two J310's connected as a DGM which are much more readily available and can be had for 20 cents a piece. The output side of the 40673 in K1BQT's design is connected to a band pass filter (yes OBTE that is a BPF) and in turn that drives the IRFD1Z3 which has a low pass filter on its output -- the classic 3 pole. Today you would probably want something a lot stiffer than 3 poles! The IRFD1Z3 is literally "unobtanium" but it looks like the BS170 would work and I will runs some tests to affirm that choice.
  3. So N6QW's Plan would be to follow the DGM stage with a switchable Band Pass Filter and using the W3NQN's stock filter design follow the BS170 stage with these filters, again switchable.  Rick, K1BQT, was careful to note that if you placed his design on other bands --you needed more filtering. To recap we will be changing the driver stage to include J310's with a BS170 amp stage and more/better and switchable filters. There is some solid sense to have a LPF follow the BZ170. In the original design the IRFD1Z3 could produce 300 MW of output thus you could make this a Qrppp rig and bypass the final stage or pick off the output and run that into a transverter for use on VHF or UHF SSB. Lots of possibilities but definitely more filtering needed.
  4. While we are on filters, when I built the LM373 transceiver which can be seen at http://www.n6qw.com  I was faced with a situation, because of this being a single IC transceiver of how to add Band Pass Filtering without using a lot of exotic switching. The answer was two Band Pass Filters with one in the receive side and the other on the transmit side. So while it adds a bit to the project in terms of cost and hardware, I am proposing two sets of band pass filters. The original design did this! You will note the BPF ahead of the MC1496 receive mixer and we just discussed the one following the 40673. The only difference we will be having switchable filters depending on the band.
  5. The original design used a 28 VDC RF device the MRF138. That is one expensive device and was a bit of a problem in that portable operation would be difficult but not impossible. One approach would be like the Sideband Engineers SBE-33 vintage 1963 --it was an AC only unit and to use it mobile they had a DC to AC inverter. Small 150 Watt DC to AC Inverters can actually be had for a $20 bill -- but that is not so easily done if you want to take it trail camping. Technology offers us netter solutions today. One solution is to use an IRF510 and limit yourself to 5 Watts. Lots of QSO's can be had with 5 watts. Other RF MOSFET's from Mitsubishi (RD series) can put out a lot more juice at 13 volts DC -- like 20 watts. There are kits being sold that offer all that you need [Check the K5BCQ website for available kits.] Communications Concepts Inc. sells a Bipolar kit that will produce 20 Watts with 100 MW of drive --another option for you.
  6. Now comes the "how to switch the filters" part. We have previously identified three locations where switching is needed actually there are 5 places. 1) the LO, 2) Rx BPF's, 3) Tx BPF, 4) Driver Tx LPF and 5) Final Amp LPF. So how do you do this? We have many options and I will take a few minutes to look at a subset of the switching. Boiled down we have basically two options for the switching and those include diode steering and relays. Diode steering relies on the principle that under certain conditions a diode when biased properly can act as a switch for AC signals. Typically these are low level signals although a company located right here where my laboratory is located (Newbury Park, CA) makes high power RF capable diode switches -- these are like watts not microwatts. But for higher power levels (and most ham budgets) a relay does the job. I purchased a stock of surplus micro-sized relays and so that will be my option.
  7. We are now circling back to our Arduino and its control capability. In several transceivers I have used switches for input and for controlling the switching of the BPF's and LPF's. In the FPM-5 I used a two pole 6 position band switch. One pole provide input to the Arduino to change bands and the second pole powered on the appropriate relays commensurate with the band chosen. In the Big Kahuna rig a simple DPDT switch did the same job. In my KWM-4 I took advantage of the K5BCQ controller that produced a 3 digit BCD code as you changed memory channels. Not knowing any better , I took the 3 digit decode and decoded it using a 74 series chip that took in BCD and the output was a digital pin that went high with an open collector. The output from the IC fed a PFET that powered on the relays in the KWM-4. Later I found it could have been done with a single IC --a CD4028 -- same principle as what I designed. So we could generate a 2 or 3 digit code -- 2  = 2^2 = 4 possibilities and 3 = 2^3 = 8 possibilities and decode those for the band switching. Now we could just take an output from an Arduino pin  and "hot up" a 2N3904 and switch the relays. The uBitx from VU2ESE takes this one step further --it reads the frequency internally generated and automatically switches in the proper relays. In the final analysis I will use either the DPDT or a real band switch. Your choice. Lots of possibilities. [For those who lurk the EMRFD and uBitx forums --the decode switching was done 5 years ago by Pete the Genius.]


 

Today's journey was to explore some of changes we will be making to the original design and what we have dubbed the IC Transceiver. Essentially the IC mainboard will remain intact.

  1. The receiver front end will have switchable Band Pass Filters.
  2. The Driver stage will use the J310's configured as a DGM and the driver will be a BS170. This stage  will have switchable Band Pass and Low Pass filters.
  3. Micro-sized relays will provide the signal switching/steering except for the diode steering on the mainboard.
  4. The Final will be an IRF510 and have switchable Low Pass Filters.
  5. The Arduino Nano & Si5351 will provide the LO and BFO Signals.
  6. Band  switching will be either a DPDT Toggle Switch or Rotary Switch.
Stay Tuned there are exciting things happening.

73's
Pete N6QW


1/11/2018 ~ More on Copper Coated Vector Board .


Firstly here is where you can purchase the Copper board. Yes it is expensive --about $28 for a piece. But if you follow the process -- you could likely get three mainboards out of a single piece. [DigiKey is the supplier.]
 
 
Now I have also done work with this type of board using Surface Mount Devices and the following two photos should expand your mind to see the possibilities. Once again Pete, The Genius is ahead of the curve. The process starts with a Fine Point Sharpie Pen and connecting the dots of the areas where you want to create an island. Next using a pencil draw a couple of guidelines around the area identified, which is then followed using a steel square and exacto knife remove the area between the guidelines. Leave at least one open hole in the island so you can connect wiring. [Yes Virginia this was done seven years ago and not invented last week by an OBTE on the EMRFD or uBitx Reflectors.]


In case you haven't guessed the hole spacing is 0.100 inch which makes it ideal for SMD application using the larger size 1208 parts.

This approach affords you the opportunity to rapidly prototype a circuit using short direct connections and a solid ground plane provide excellent RF properties. The other distinct advantage is that should you need to shield circuits a piece of scrap PC Board can be soldered directly to the mainboard. I have found it necessary at times to provide additional shielding and it was a simple matter to simply solder the shields in place. 
 
Stay tuned as we quickly move along with these projects -- you need to start drinking energy drinks and lots of coffee to keep up.
 
73's
Pete N6QW
 

1/10/2018 ~ More Examples of Single Sided Copper Vector Board used in my Rigs.

Keep in mind short direct connections and a common ground plane are always the "best" approach when building RF circuits. Tell me when it is not? There are probably one or two OBTE who will tell you they heard something third hand on the EMRFD reflector that it was not --my response consider the source.
 
In 2007 (that was 11 years ago) I started building transceivers using a common IF frequency ~ 4.9152 MHz and at the same time was actively pursuing an elusive goal of a shirt pocket sized SSB transceiver. I have come close to getting to the shirt pocket SSB goal; but not the final goal --one built entirely in an Altoids tin --maybe in 2018?
 
One of the 1st attempts was a 17 Meter SSB transceiver and like many of my projects I built two versions --the first typically looks like crap and is a true prototype. The second builds usually incorporate the "ahh" moments where you realize the better way. The K1BQT 2nd build will take advantage of what I learned from the 1st build.
 
The 17M  transceiver used a couple of innovative approaches. First using the 4.9152 MHz IF places the LO above the incoming at around 23 MHz. For the OBTE -- 23.04 - 4.9152 = 18.1248 MHz which is of course right in the middle of the phone band on 17 Meters. So how does one get to 23.04 MHz. The main frequency control is a VXO (variable Crystal Oscillator). Using several crystals in parallel configured as a Super VXO it is possible to move the frequency 10 to 30 KHz and thus you have a VFO like action using crystals.
 
The icing on the cake is the number of cheap computer crystal available to us. In this case, one of the stock frequencies is 11.520 MHz. The OBTE will quickly say --hey that is not 23 MHz. But if you use a diode frequency doubler (Thanks to W7ZOI) you can turn that 11.520 into 23.04 MHz. Now the real beauty is that what is ever changed in frequency at 11.52 MHz is doubled. So a 10 KHz swing at 11.52 MHz is now a 20 KHz swing at the output of the doubler circuit.
 
But I soon found that I did not have full coverage of 17 Meters. Thus a second set of crystals was custom ordered (ouch $50) and using a small relay I could switch the bank of crystals used in the VXO. Now with four crystals I can cover almost all of 17 Meters. I am missing a 6 KHz of coverage from about 18.141 to 18.147 MHz. Keep in mind this pre-dates the Si5351 and now $20 would give you full coverage with a Color Display -- but again this was 11 years ago. I may yet retrofit this rig --maybe with a 1/2 size OLED.
 



 
The above photos show how the single sided board was used for the 17 Meter  transceiver. Two things came from this 17 Meter project which were used about 4 years later in the shirt pocket transceiver on 20 Meters. The two items are the single sided copper vector board and the crystal switched VXO. Actually I should also include --the same IF frequency 4.9152 MHz was used in both rigs! The 20 Meter Shirt Pocket Rig photos (V1 and V2) are shown below.
 




Again the real advantages with the single sided board are short direct connections and the common ground plane. Today we have seen two examples of transceivers that were built using the single sided copper vector board. Also don't discount the crystal switched VXO as a LO source and the choice of 4.9152 MHz for a homebrew Crystal Filter. [If this choice of IF was good enough for Elecraft in the K2 why not in your rig.]
 
I sense the excitement building --hang in there as what I am covering now will be very useful if you decide to build the IC Transceiver. Some other parts you need to start finding --aluminum spacers with 4-40 threads. Look for 1/4 inch and 1/2 inch high spacers. The initial size goal for the mainboard is 4.5 X 6 inches. I will know more once I go through the iterative process of parts layout.

BTW in case you would like to see the documentation for the 17 Meter Transceiver see the link below:
 
 
Yes Virginia, in addition to there being a Santa Claus there are hams scratch building real homebrew transceivers and the 17 Meter/ 20 Meter projects are original designs. So what are you waiting for --get off that couch and start building!
 
73's
Pete N6QW

 

1/09/2018 ~ Alternate Part Sources for the IC Transceiver. Also Tribal Knowledge on the Construction Practices to Build the Rig.


I love it when someone can find better parts bargains for projects and so it is with my friend Bob, N7SUR. He advised me that the MC1350 @ $2.50 and the MC1496 @ $1.35 can be purchased from Dan's Small Parts in Missoula, Montana. Dan can be found via Google and indeed has a lot of hard to find parts and some amazing prices.
 
Several years ago I purchased some vertical Style S Meters from Dan and these can be seen on my Belthorn III and Big Kahuna SSB Transceivers. The cost was very reasonable. May need to look to see if he still has those meters.

About 10 years ago I did build the K1BQT transceiver for 20 Meters and was delighted with its performance. That rig was given to another ham and now my desire to build a second unit. But this one will be a departure from my extensive use of the CNC Mill as it will use a technique that is quite excellent and not Manhattan nor Isolated Pads. The construction by the way is probably as fast or faster than having a specially fabricated circuit board.
 
Vector (as in Vector Board) makes a single sided copper Vector board and comes in sheets like 4.5 inches by 17 inches long. It is not cheap but one piece is enough for several projects. The beauty of this board is that the top side is a common ground plane. So any connections to ground are simply soldered to the top of the board. Below is the W7ZOI HYCAS IF amplifier strip/Product Detector  as built for my JABOM Transceiver project. Once it was built and tested I soldered a copper strips around the sides to form a copper box enclosure. There was a trick here. The finished board was mounted on aluminum spacers during the soldering of the copper sides. Now when you are done the whole assembly can be screwed down to the chassis base plate. The second photo shows how the point to point wiring was done on the under side (insulated) of the single side board. Those who like to pick things apart (as they do on the EMRFD and uBitx Reflectors) there is only one connection that crosses over the wiring. [Further note -- this project was built 7 years ago and described in a QRP Quarterly Article. Pete the Genius was once again breaking new ground!]
 


 
 
Now I am going to share some Tribal Knowledge about how to make the part layout and minimize cross overs and have the parts laid out to minimize connections. In the case of the HYCAS it was an iterative (meaning doing several times) process that first starts with a piece of PLAIN (no copper) vector board and the schematic that you will use. Liberal use of small wood blocks and masking tape facilitates the process. Another tool is a metal cookie baking pan that is about 12 X 18 inches. Typically these pans only have about a 1/2 inch high side.
 
Start first by taping the wood blocks to the cookie pan base spaced slightly shorter than your plain vector board. Next tape the board to the wood blocks  and assuming the builder knows where is pin 1 start by inserting one of the IC sockets (yes 14 pin DIP Machined Pin Sockets, MPS) into the board. Using the schematic insert parts into the board so that you get a compact layout and visually see how to minimize connections and cross overs. Take small sections at a time to do this. When you are satisfied that you have all of the parts on the board for that section and that the connections are minimized and the shortest possible connections are used. Take a photo with you phone camera. You will use this photo to do the final layout on the copper board.
 
  1. Now as to the why use the plain board for the initial layout --it is the oil and acid on your hands. The trick is to minimize the physical handling of the final board to minimize any discoloration and difficulty soldering based on corrosion build up. You might consider wearing rubber surgical gloves during the final soldering process. I do.
  2. The plain board with the cookie pan enables you to stop work without having to clear off everything and just moving the pan out of the way keeps everything neat and tidy. It also helps prevent the loss of parts as the pan acts as a captive mechanism.
  3. The process really helps you understand the circuit elements as it goes beyond just stuffing a board with parts like the OBTE frequently experience without the knowing the why.
  4. The wood block are about 1 inch high which lets you push through the leaded components so that they are flush with the plain vector board. The blocks are a must!!!!
An important tool must be built prior to working on the final single sided copper board and this is shown below and is comprised of two parts. A 1/8 inch knob and a 1/8 inch drill bit. Resist simply taking a 1/4 inch knob and a 1/4 inch drill bit like the OBTE would do. This tool is used to remove a small amount of the copper material around a part that will be placed on the board and not grounded, meaning it will be connected to other components on the underside of the board. Before you insert any non-grounded component, twist the drill bit a couple of times to remove the copper around the penetration hole and then inspect the whole to assure no burrs remain.

 
The reason the 1/4 inch should not be used is that too much copper would be removed. You want to remove just enough so the parts are not shorted to ground but not too much to reduce the ground plane effect. In the case of the IC sockets I remove copper for every pin (14 or 8) and you will also find since I use machined pin sockets (a must) that each hole for the DIP sockets must be slightly enlarged to accept the machined pin sockets. If you don't know what a machined pin socket is --turn off your soldering iron and take up another hobby. The why of the MPS is to assure the socket doesn't float around and that you have a solid pin to affix components to and finally a solid soldering base.
 
 
 
 
 
This is a good place to stop --today we gave you an alternate source for the IC's, introduced you to the single side copper PC Board method of construction and shown a special tool you will need to build the project.
 
73's
Pete, N6QW

 

1/08/2018 ~ Getting a "Head Start" on the K1BQT Rig


So OK you got some Amazon Gift Cards for Christmas and are wondering where to spend the loot. The very 1st thing you do is go to Bill's, N2CQR, SolderSmoke blog and click on the Amazon Link and begin your shopping for some of the parts to build this rig.
 
I just know there are many homebrewer's who simply can't wait to build this amazing rig --even before me. So if you have that kind of itch to scratch, then visit my website http://www.n6qw.com 
 
On that site you will find an amazing array of projects from the N6QW Laboratories and only one or two even hint of being Bitx related.
 
The project you want to view in detail is "The Big Kahuna" which is extensively documented. There are three pieces from that project that will be used with the IC Transceiver (That is now what I am calling K1BQT;'s transceiver.)
 
The three pieces are:
  1. The Arduino Sketch for the Big Kahuna uses the 9.0 MHz IF and thus already puts you in the ball park. This sketch has you use a DPDT toggle Switch for band change with 1/2 the switch providing the input to the Arduino to change bands and the other half provides voltage to the relay banks to select the proper set of matching Band Pass and Low Pass Filters in sync with band showing up on the display. The display is the very large 320 X 240 and offers some possibilities for displaying other data.
  2. Now Genius Guy that I am --the sketch includes the code for 5 Bands which are linked to five pins on the Arduino -- thus by selecting the proper pins you can steer your build to what ever two bands you chose. Like 80/20 or 80/40 or 20/15. Guys with a 9 MHz IF I have chosen to stay away with anything saying 17 Meters. Now if you change the Filter Frequency say to 8 MHz then you can use 17M; BUT you will have to change the numerical data in the code to account for the IF offset and the new BFO frequencies. You can do that --don't ask me to do it. The wiring of the display and the level shifter (CD4050) is embedded in the sketch.
  3. The Band Pass Filter Data for the 20/40 selection is in the Big Kahuna complete with LT Spice Simulation, plots of the curves and the component values. If you chose other bands you will need to develop the schematics -- the LT Spice schematics should be a clue for you. Don't ask me to do it
  4. The Low Pass Filter Data is a lift from W3NQN so you have the numerical data for the various ham bands. The calculation of the turns for other bands is your exercise not mine. This is a good time to venture out on your own.
So having these three pieces already available eliminates about 35% of the project that will need development. We will endeavor to use other N6QW building blocks such as the driver stage [EMRFD lift] and the IRF510. Thus even more hardware can be built ahead of the main IC board.
 
73's
Pete N6QW

1/07/2018 ~ A Link to the 1985 HR Issue

Thanks to another Pete in very cold Illinois he found a link to the issue where the original article appeared.

http://cahabatechnology.com/aa4pg/hrmag/hamradiomag/Ham%20Radio%20Magazine%201985%20%20Issues%201-12/11%20November%201985.pdf

Thanks Pete!

73's
Pete N6QW

1/06/2018 ~ Coming Up in the Queue - Start Planning Your Moves and Locating Parts! (See Block Diagram Add)


 
K1BQT's XCVR from November 1985 ~ With 2018 Updates!
 
 
 
Visit Jameco Electronics http://www.jameco.com and you can find the following critical parts
 
  • MC1350 IF Chip P/N 24942 Price $3.95
  • MC1496 (four required) Double Balanced Mixer P/N 23211 Price $1.95
  • IRFD1Z3 (not available anywhere) But I believe a BS170 will work
For the IF Filter (9.0 MHz) I will be using the Crystal Filter from the GQRP club. I believe they still have a very small quantity left in the bins. You will also need to find some 40673 Dual Gate MOSFETs or you can use two J310's configured like a DGM. The Final Amp can be a IRF510 (also at Jameco). This will be a two band radio --a simple toggle switch for your two bands of choice like 80/20 Meters or 40/20 Meters. Guys forget 17 Meters with a 9 MHz IF. My design will have separate band pass filters for the two bands.
 
K1BQT's design has a key feature (maybe that is where I got the idea for my last 2018 project) and that is the MC1350 IF amp chip is used on both transmit and receive. Also his rig not only had AGC but ALC -- just a few more parts. Like in my last rig --it will have a cool analog S Meter OR the S meter included on the Color TFT.
 
The updating will include the Arduino Nano, Si5351 and the 160X128 Color TFT display.
 
This is not the very next rig on the bench; but the info is being provided early so you can start to get organized. The first trick is finding the article. The Ham Radio Magazines were free on line. But someone evidently filed a law suit and they were removed.

Found: http://cahabatechnology.com/aa4pg/hrmag/hamradiomag/Ham%20Radio%20Magazine%201985%20%20Issues%201-12/11%20November%201985.pdf

I will create schematics of what I build since it is different than the original article. This should not violate any copyright issues. BTW the original article used the TO-5 Version of the MC1496 and the ones being sold are the 14 Pin DIP --not to worry I will provide the magic decoder ring.

Below is a block diagram of how this new transceiver (new for 2018) might be fashioned. The building blocks are pretty standard.


A Post Script --the Ham Radio Magazine cost but $2.50 and unlike the ham radio magazines of today did not focus on contests and operating news -- this was a publication for homebrewer's. (For the box top extra's this means something other than appliance operating your latest ICOM or Yaesu)

Post Post Script: On eBay currently is a IF board out of a commercial radio (being sold from Israel) that has three 9 MHz Filters on the board. I had an earlier posting about a rig now in work that uses two of the filters (USB/LSB). That rig is still in the queue for 2018. The board from one of the sellers costs $34 including the shipping . But also on the board are the critical MC1350 and several MC1496's. So this board already contains many of the parts. NOW a huge caution if you are unskilled with a soldering iron and the removal of parts from a manufactured board --this is not a good option for you for the parts other than the filters.
 
73's
Pete N6QW

1/05/2018 ~ Listening to the new Rig!

 
The 1st SSB Transceiver of 2018 from N6QW ~ 35 years in the Making!
 
 
 
 
It only took 35 years but we now have achieved our goal. High marks for an Old/New Technology Rig. Get off that couch and start building --we soon will moving forward on another new rig --or I should say and old rig that has been reformed! You got to keep up.
 
73's
Pete N6QW

1/03/2018 ~ It is alive!

I am happy to report the 1st QSO with the reformed rig and it was with a station in Hawaii, Thanks Norm for the 1/2 hour QSO. (An oddity in our 5X9 running a FLEX 6700 rapid fire QSO Mode.)


The D-104 microphone gives you an idea of the size. I was also constrained by what holes in the front and back panels existed as a result of the original build and several reincarnations. The upgraded changes include USB/LSB select, 128X128 Color TFT, a new driver stage (2N3904/BD139) and the IRF510 as a final amp. the IF is at 9.0 MHz.
 
Some changes to the main board include removing the homebrew diode ring single balanced mixer that was used as the Product Detector on receive and as the transmit mixer stage on transmit. A TUF-1 replaced all of that hardware. I also removed the homebrew diode ring balanced modulator and replaced that with an ADE-1. Man the new devices sure replaces a lot of old hardware.
 
I mentioned in an earlier post one of the problems I found with the original build -- a bad diode in the single balanced mixer used as the PD / Transmit mixer --this is what prompted the shift to the TUF-1 and while I was at it I changed out the original balanced modulator to the ADE-1. The TUF-1 is just below the GQRP Crystal Filter and the ADE-1 is mounted on a vertical board just behind the panel meter. You will see some open space behind the vertical board --all of the was formerly jam packed with Balanced Modulator components. The microphone amp was a 741 ( a staple for 1980's transceivers). Just behind the vertical balanced modulator is a brass shielded enclosure that originally housed the BFO. In that now resides a relay that switches power to the receive and transmit circuits that do not run continuously. BTW I bought these relays which have 12 VDC coils and are SPDT capable of handing 8 amps --from Jameco Eeletronics on closeout -- 10 for $6
 
 
Mind you again this was originally built in the 1980's and the architecture relied heavily on diode switching. On the receive side the 1st stage is a Dual Gate MOSFET mixer which was diode steered through the IF chain. The other half of the diode steering was the Balanced Modulator.
 
The IF chain used a pair of 2N3904's on either side of the 9.0 MHz Filter similar to what ON5FE used in his 1970's Transceiver IF Module. That stage is followed by a Motorola MC1350. This device was chosen so I could add AGC on receive and an S Meter. On transmit this stage has fixed gain applied. Following the MC1350 is the TUF-1 where on receive the BFO is diode steered and we have product detection where from there it goes on to the audio amp stage (2N3904 / LM386-3). On Transmit the LO is diode steered to the TUF-1 and this now is the transmit mixer.
 
Following the TUF-1 (on transmit) is a 20M Band Pass Filter and then on to a 2N5179 RF amp. The 2N5179 was the Hayward device of choice in many of his SSDRA circuits. From the mainboard we have the driver stage and the final amplifier.
 
One challenge I had was space -- the box was already pre-determined! Where was I to mount the TR relay and the Low Pass Filter. My mechanical engineer son (the one with the super knack) has decoded what I want for Birthdays and Christmas presents. Yes, Hardware! Thanks to son Nick I am well stocked on nuts, bolts, spacers and connectors! Here is how I resolved the TR relay and external amplifier switching. I used four 1 inch standoffs to mount the small circuit board to the same mounting holes as the copax connector. Right near that is the small circuit board built on the CNC that is the LPF. The connections are short and the assemblies fit in the cramped quarters
 
 
Below are a couple of shots of the Final Amp board (IRF510). I have designed a standard RF amp Circuit board that can accommodate most any RF device that comes in the TO-220 package style. Thus some of the island squares are not used for the IRF510. When I need a new board --chuck up a piece of stock in the CNC and press "Start".
 

 
I will continue to fine tune this rig as there are some functions the need to be completed. I have a small push button on the front panel that will initiate the Tune Tone. That wiring needs to be installed as well as build the three stage RC filter to take the square wave 988 Hz tone generated by the Arduino and convert that to something close to a sine wave which is then routed to the balanced modulator for Tune Up.
 
73's
Pete N6QW

Well these are only new if you consider that they will be/are newly reformed old projects.


I often mention that I have two boxes of projects ~ one very large box with ones that sorta worked, worked once and died, never worked and finally never would work as configured. The other much smaller box are ones that work --and nicely I might add. So I have dug really deep into the very large box and pulled out three that in their present state are not air worthy; but the goal is to make them so. Along the way we will add goodies like digital VFO's, Color TFT display and only commercial filters. Frankly I have better things to do than screw around with Dishal software.
 
The first which I have already started is about 75% there and shown below. This was a solid state rig built in the early 80's and while it did work --was never quite what I thought is should be. Recently I found out why. I will cover that in a future posting.
 
By the way most of the stuff that is in the large box was built using the SSDRA [Solid State Design for the Radio Amateur]. Do what you can but get yourself a copy even if that means selling your EMRFD. I just think for someone who would like to start building radio projects EMRFD is not the 1st choice and SSDRA will get you farther and faster along the learning curve. Just my opinion but there is a sound basis for that opinion. I have found a good use for my EMRFD --it is a great bookend!
 
 
The above rig started life as a 20 Meter QRP rig with an analog VFO, Crystal BFO, 9 MHz KVG Filter and an IRF510 Final. This was not someone else's design but something I conjured up using building blocks taken from various publications and reference books. This rig was built Pre-Internet. Many of the circuit blocks are from Hayward and DeMaw and some pulled out of thin air.
 
I will document what I did and how the radio was constructed --for now just drool over the photo.
 
BTW I am now in therapy after participating briefly in the Bitx20.io forum. I have since sworn off reading any more posts from that group. The uBitx is a superb rig and I will stop there!
 
Happy New Year!
 
73's
Pete N6QW

 

15 comments:

  1. I'm a droolin'!

    73 & HNY! Steve N8NM

    ReplyDelete
  2. "BTW I am now in therapy after participating briefly in the Bitx20.io forum"

    LOL, you are not alone in that, it took me a few weeks to get over it and I know from other forums that many others felt the same way also. For an outsider it can be a not to pleasant experience.

    Looking forward to reading what you have been up to Pete and getting back into building some things myself. Happy New Year.

    Rob.

    ReplyDelete
    Replies
    1. Hi Rob,

      Regrettably some of those posting on certain forums have solder burns on the palms of their hands from grabbing the soldering iron by the wrong end. Some even hold the highest class license here in the US. I hope to update the blog today -- another rig on the air!

      As always nice to hear from you.

      Pete N6QW

      Delete
  3. Replies
    1. Hi Dean,

      Thanks for your kind words -- just glad I got it working.

      73's
      Pete

      Delete
  4. Stunning, Pete! The picture of it alongside the D104 really puts into perspective how compact the rig is... It's an amazing form factor for a homebrew SSB transceiver from the early 80s - Wow! Good stuff!

    ReplyDelete
    Replies
    1. Hi Steve,

      It only took about 35 years to get all of the kinks worked out. This has been a fun project in that a single device can replace many components. Worls better too.

      Pete

      Delete
    2. You were just ahead of your time and it took 35 years for the technology to catch up!

      Delete
  5. Pete, If you keep talking about making your shirt pocket transceiver smaller, I will be forced to go smaller too. The NE602 rig okay, but I think I can do better!! I look forward to more of your builds.....

    Dean AC9JQ

    ReplyDelete
  6. The vector board technique looks very interesting. I might have to give it a try..

    ReplyDelete
    Replies
    1. Hi Dean,

      I got an tip that there is a listing on eBay for a piece of the board that is 2X the size of that listed in the link --for about the same price as the board in the link. 2 for 1 is usually a good deal; BUT Caveat Emptor

      73's
      Pete

      Delete
  7. Pete, Thanks for championing K1BQT's superhet SSB design of 1985. HR was available to me as a youngster and even though a lot of it was beyond me back then I read every copy I could get with great interest. We regarded it the Rolls Royce of builders mags down here in my VK family. This particular rig looks like a good design, I always got my MC1496 bal modulators and mixers to work, and I have a bunch of SOIC 1496es so who knows, this one might be in the queue for me. Watching with interest.

    Most interested to see how you will band switch the rx and tx signal paths, there are 3 points ... rx BPF, post tx mixer BPF (I would strengthen that filter over the original design) and the LPF. Relays? Or a mix of diode and relay switching?

    I also like the custom 28v PSU, for 20+ watts from the IRF510. It will run perfectly well on a 12-15v LiPo pack, with the only difference being 5 watts. Sweet!

    73 Paul VK3HN.

    ReplyDelete
    Replies
    1. Hi Paul,

      Many thanks for your post. The questions you posed are being worked and will be answered on the blog. The plan is to mostly use the IC part of the design and make the necessary changes to the topology so that it will be multiband. The block diagram posted several outlines some of the proposed changes.

      Thanks again -- watch those solder burns.

      73's
      Pete N6QW

      Delete
  8. OK Pete you've answered my questions. I've used mini surface mount relays for filter switching and they work a treat, tiny, DPDT, low current, available in 5v or 12v varieties. The Omron o es are premium.

    I have typically paralleled the two sets of contacts with one relay at each end of the filter for best input-output isolation. But this doubles the number of relays you need to buy and place on the board. What do you plan to do in the small signal filters, one DPDT relay or two? Paul VK3HN

    ReplyDelete
    Replies
    1. Hi Paul,

      Up until the source dried up I was using the Omron G5V12 series and they were ideal. These were the SPDT variety and they worked exceptionally well. Many of my rigs have these. Mostly I use SPDT and have found some cheap replacements that I am getting for 60 cents a piece. These have flimsy contacts so I mount them upside down with superglue. If you check the Simpleceiver Plus SSB posts on the blog you will spot them. I buy them at a place called All Electronics in Los Angeles.

      Since this will be a two band rig the SPDT will work nicely for the various circuits.

      73's
      Pete N6QW

      Delete