11/23/2018 Link to Arduino Sketch

I would like to take a moment to recognize one of our visionary hams ~ Herb Johnson, W6QKI (SK) for his pioneering work in advancing the state of the art in our ham radio equipment. For those who don't know, W6QKI was the founder of both Swan Electronics and Atlas Radio. Many of those long in the tooth Swan and Atlas Radios are in daily operation to this day. I have an operational SWAN 120 SSB Transceiver built in the 1963 time period which has recently been used on 20 Meters! Thanks Herb, for seeing something more to our wonderful hobby than just a one tube crystal controlled transmitter and two tube regen receiver. There is an exceptionally warm spot in my heart for Herb -- we both find our early roots in Southern Western Pennsylvania.

In passing it is also interesting to note that the foundation of our amazing hobby started with that Italian guy --you know Marconi. Then there was E H Armstrong the Father of the Superhet and FM Radio! In time, the hotbed of ham radio equipment development passed on to the US with names like Art Collins, Bill Halligan, Jim Millen, Al Kahn, Bob Drake, Leo Myerson, Edgar Johnson and Faust Gonsett leading the charge. Just look at any shack photos from the 1940's-1970's and you will see their equipment on the operating desks.

Then the wave shifted to the Far East (ICOM, Kenwood and YAESU) and those are the common names today. But lest we forget there is still leading edge ham gear made in the USA today thanks to FLEX (Youngblood), Elecraft (Swartz and Burdick) and Ten Tec (Kahn is a SK but the compnay is still building ham equipment). There is also a new wave in low cost kit type radios that largely through the Internet have a huge following and those include Ashar Farhan (Bitx), Hans Summers (QCX and soon QSX) and Tim Walford (Upton Bridge Farms).

We have many to thank for the gear we now use in our shacks. Then of course there are a number of hams who homebrew their entire stations; but that group is small by comparison and not well known. That said, if indeed you are a homebrewer, today is an amazing time for us as the availability of new low cost, easily deployable technology is staggering. How about a complete Digital VFO with Color TFT display for a $20 bill? In the pure homebrew arena there are many technology giants such as Wes Hayward, W7ZOI, Doug DeMaw, W1FB and of course Rick Campbell, KK7B and many others who are too numerous to mention.

A Long Way Around Journey!

Several years ago I was interviewed by Eric Guth, 4Z1UG in his podcast QSO TODAY (#60). Over the course of the podcast we talked about how that now many current introductory ham rigs "start" at about $3000 and go on up from there. Although the ICOM IC7300 can be found NIB at slightly more than $1000 -- probably the reason why there are so many now in use. During our discussion I suggested that a ham could buy an older rig on eBay for about $100 and with a little "digital magic" have a very capable modern rig. Sure it would lack all of the menus and audio enhancing do-dads. But it would do three important tasks: hear well, transmit well and have accurate plus highly stable frequency control. The station at the other end can look all day long at your signal on their 72 inch LCD screen attached to their FLEX 6700 and not make bad comments about the RF coming out of your rig. Other than perhaps the disparging comment that usually results from the FLEX operator in his statement "Well it looks and sounds OK for a $100 rig". Keep in mind he has to constantly justify why he spent $10K on a ham rig and here you are with $100 make over looking just FB!

With that as a background I am constantly looking for the $100 priced older boat anchor rigs that can be reworked to provide a very capable and good sounding rigs that are useable on today's hambands. Of course having digital mode capability is a must which totally drives frequency accuracy and stability. Finding that $100 boat anchor is getting harder to do as the the older rigs have gone up in value based on the price points of the new starter rigs. But bargains can be found!

Some of my $100 past projects have included upgrading the analog version of the Ten Tec Trition IV (Model 540) to the digital version (Model 544) and lest we forget the LCD Digital Display/Frequency Stabilized VFO embedded in a Heathkit HW-101. Then there was the conversion of the Ten Tec Model 150A, a commercial crystal controlled rig that I converted over to a ham rig with LCD Display and a Si5351 VFO. Add to that list the hallicrafters FPM300 (2nd Row) that by using just the solid state mainboard from thi boatanchor I completely repackaged the unit into a five band QRP rig . This an extreme makeover but is in use today.

[Just a footnote here... a supposedly untouched, operational FPM300 is currently listed for sale on eBay at a starting price of $1500. Good Luck with that buy... even when they worked, they were marginal on frequency drift and mechanical stability not to mention self oscillation under some load conditions. A person who worked on the line building the FPM300 saw my you tube video of the completed unit and he shared with me that my rework sounded better than anything he ever saw coming off of the line.]

The Atlas RX110/TX110 Twins

you tube video Atlas RX110/TX110 with the Si5351

Pictured at the top of this piece is the Atlas Radio RX110 and TX110 Transreceiver. Basically this was a later offering from Atlas Radio where you could buy a fairly rudimentary 5 band receiver as a starting point. The next step was to add on a small box that housed a QRP transmitter section whereby using mostly the receiver circuitry you instantly had a no frills bare bones five band, 10 watt, QRP SSB transceiver. You can checkout the Internet for reviews of this rig. VK3YE, Peter Parker has a review in eHam in 2003 on this pair --Peter also very deliberately used the word "rudimentary" to describe the rig and the link to his review. https://www.eham.net/reviews/detail/3606 There is reference made to call these the "Atlas Twins" much like the Drake Twins. But in reality it is more like the R4 Receiver and T4 Reciter. Although the R4 and Rx110 are not in the same league as to capability, functions and features! Keep in mind RX110 = No Frills! Thus my latest project will involve the "Atlas Twins"!

• Link to Transmitter Repairs
• Link to Receiver Repairs
• Link to Arduino Si5351 Interface

For those who don't want to read on and simply like to push buttons and do no thinking or evaluations click on the above two links. For those who want to understand the process read on.

Presented above is the RX110 Block Diagram. We can see some important information in the diagram. Starting on the left side, PC050 is a board containing a series of bandswitched Band Pass Filters which are selected by the same bandswitch that controls the VFO range. The signal after passing through the BPF is fed (to PC-110) to a 1N914 Diode Ring Mixer which is followed by a "strong" post mixer amplifier --the 2N3866. For those who have SSDRA from Wes Hayward, W7ZOI he explains the why of such a Post Mixer Amplifier circuit. After passing through the PC-110 board we now have a resultant signal at the IF Frequency of 5.595 MHz.

Next in the path is the PC-310 board which has the SSB Filter, IF AMP (MC1350), and the Product Detector, a venerable 40673 (DGM). An LM3900 provides the Audio Pre-amplifier and AGC control with a very healthy LM380 for the Audio Output. Also on the board is the BFO comprised of a single MPS6514 oscillator stage. Again go to your SSDRA and you can see circuits by W7ZOI for the MC1350 IF Amp and 40673 Product Detector. Forget EMRFD you will be lost!

Did you catch that there is NO RF amplifier stage which follows the typical Atlas architecture. Thus the receiver should be quiet and not prone to overloading. VK3YE mentions the low internal noise in his eHam evaluation.

In looking over the the "devices used" it is obvious the idea was to provide basic circuit functions using low cost parts. Double balanced mixers using 1N914 diodes and common transistors such as the MPS6514 (a favorite of Ten Tec) keep costs down but provide good performance. Drag in a MC1350 IC and you have an AGC controlled IF amp strip. There were no heterodyne crystals to move the signal to other bands. Instead the VFO was shifted in frequency to work with a 5.595 MHz crystal filter. Imagine a stable analog VFO operating directly at 24 MHz to enable USB operation on 10 Meters.

I shudder to think ( because of what we now have available to us today) of trying to build a stable analog 24 MHz VFO. Parker mentions frequency drift and the inability to precisely set or return to specific frequencies. Much like the other Atlas transceivers the analog VFO display was a circular drum assembly mounted on the top portion of the VFO compartment. Using an ingenious mechanical system, as you tuned the VFO knob the Display Drum rotated thereby exposing a small portion to the dial face window. A VFO frequency table in the manual shows that there was only one BFO crystal used so that the LO was above the incoming frequency on 80 and 40 Meters which resulted in a sideband inversion (LSB resulted). Yet on the higher bands the LO was below the operating frequency so that there was no sideband inversion and boom USB. At that time this was OK; but today if you want to operate the digital modes like 40 Meter WSPR or FT8 you need USB. [Did I mention that the analog VFO would be marginal with the digital modes!] Thus one BFO Crytsal covered the normal modes of operation for the 80-10 Ham Bands. That reduced cost and saved on parts and labor to install the parts!

But the block diagram as at a high level and there are actually more devices in the radio than are shown. For instances the BFO is two transistors with the second MPS6514 being a buffer stage. So the Block Diagram gets you in the Ball Park and the schematic gives the Section, Row and Seat Number.

Believe it or not there were two variants of the transmitter section, with the standard being a QRP transmitter rated at 10 Watts across 80-10 Meters. A second variant had a pair of MRF454's to provide a 100 watt output capable transmitted signal. Essentially some RCA jacks internal tothe transmitter were either jumpered for the QRP mode or the external amp board was inserted inside of the rig and the wiring routed to put the larger amp stage in line. If the amp board was installed later on as opposd to a fcatory install then several mods were made to the low level driver board. These include removing one capactor and doubling up on the size of a feedback resistor. The resistor change add more feedback and reduced the gain.

Based on the photos on the eBay listing I believe mine is the QRP version that will be coming next week. It has since arrived and evidently did have the power amp board installed but has since been renoved. That is just FB with me as replacing the final will be a lot less expensive than a matched set of MRF454's and I do have outboard linear amps that will take that signal to over 600 watts.

One must consider that Atlas Radio initially was on the leading edge of providing compact solid state rigs. I have fond memories of my Atlas 180 installed in my Chevy Vega Station Wagon and happily mobiling to work along the Southern California Pacific Coast Highway. Using 20 Meters mobile I was working the world. Keep in mind the original Atlas Radios used licensed circuits from SouthCom International ably led by a another Ham Radio giant -- Les Earnshaw, ZL1AAX. These same circuits were used in radios he designed for the US Military.

During the 1970's I worked for Douglas Aircraft and a close ham friend was one of our company pilots, who was on assignment with Japan Airlines flying with their pilots as they were introduced to the DC-10. I had several QSO's with my friend as he was Aeronautical Mobile using a Collins 618T and the blade antenna on the DC-10, while I mobiled to work with my Atlas 180. Now that was pure ham radio.

[Author's note: there was only one thing better than mobiling to work with the Atlas 180! Traveling to work at the same time each day you often see the same cars along the road. One car in particular was a champagne colored convertible Mercedes 450SL being driven by a really hot looking blonde complete with a pony tail flapping in the breeze and of course the oversized California Sun Glasses. But that is not what was most memorable. It was the vanity license plate (no it wasn't a ham call sign) that read "SETLMNT". I could only imagine some poor bastard now driving a Volkwagen Beetle as he made the monthly alimony payments. The things we remember.]

But then the Japanese took the lead in the market by adding more features and functionality to their rigs at prices far less than just the parts in the Atlas rigs . A sure sign of low wage competition and how that can destroy US Manufacturing. It is reasoned that not many of the RX110/TX110 were actually built and sold. So this rig is not as recognizeable as the Atlas 210X where over 14000 were produced which pales by comparison to the estimated 250,000 ICOM 7300's which are now at operating positions around the world.

Thus when I spotted an operating RX110 and an inoperative TX110 combination being sold for just a bit more than $100 I made a bid and now our journey starts. The objective of this project is to turn the proverbial "sows ear" into a "silk purse".

ATLAS Radio RX110 / TX110 Rebirth

This is an example of shoemaker doing electronics work. That Kluge of wires is the first Microphone Amp stage. A former owner must have thought there was a problem and indiscriminatly cut wires and tack soldered components "wherever". One of the things that needs to be changed! It is a high gain stage using a MPSA12 Darlington Amplifier.

As usual with most of my projects the very last thing I do is to heat up the soldering iron; but the very first thing is to do a bit of "noodling" (contemplative thought process) about what and how I would approach the project. I always try to start with some goals and this project is no exception

1. The very first out of the box goal is to study the manuals for the RX110 and TX110! Read The Manual as then and only then can you understand the circuit functions and the design intent. Plowing ahead without this critical step is begging for disaster. On my soujourn through the manuals I can see how the design approach was imbued with "value engineering" which liberally translated means --getting away with the minimal possible hardware yet have a real rig. The no frills design is a testament to that statement.
2. The second goal is to make an assessment of the hardware WITHOUT any power applied. The eBay listing indicated that the receiver (RX110) was operational but the transmitter (TX110) was not. I need to verify how the circuitry physically looks. Has someone had the magic screwdrier or magic soldering iron at work here. Look for non-factory soldering or components just tacked into the circuit. Look for any shorts, burnt resisitors or leaking caps. Thankfully if the receiver works then 80% of the transmnitter works. I suspect the final (or finals if the high power version) may be blown. But find out first as a replacement RF device is about $25 or a matched pair about $60. So one needs to do some trouble shooting (much later on) before calling RF Parts and ordering a device. The final (or finals) may be OK and it is some other problem and thus you just blew some cash you didn't need to spend. RF Parts is one day away by UPS from me, so I can delay ordering any costly parts until absolutely needed. The five photos above clearly show some non-professional hack has been into this radio!
3. In looking over the receiver device list I have everyone of them and so am not concerned about blown devices or the avaialbility of replacements. Should the SSB filter be bad then I can readily substitue other filters and just adjust the Arduino sketch to handle the new frequency scheme. Again until Steps 1 and 2 are fully completed DO NOT APPLY ANY POWER TO THE BOXES! When you finally do apply power start with the receiver and get that working -- it is 80% of the total assembly!
4. The Basic Plan is to get the rig working like it came from the factory and then and only then consider the upgrades and modifications. You have to have a baseline for making any changes. Fools rush in and then it is a problem!
5. My mentioning of the Arduino sketch should immediately draw attention to the fact that I will be replacing the analog VFO and the single BFO crystal with a Arduino/Si5351. So this is one of the early goals beyond making an assessment of the rig itself. Fortunately armed with the information I already have, a majority of the Arduino development work can be done without phyically having the two boxes in front of me. The manuals tell me what I need to know such as the IF Frequency, BFO Frequency, and the frequency ranges of the five bands that are covered. This is where the Arduino shines -- just a bit of reprogramming from other prior projects and I am there.
6. The Arduino/Si5351 will now add the dimension of accurate and stable frequency generation as well as USB/LSB capability for the digital modes. My basic digital VFO/BFO/Color TFT module can be had for about $20. Such an application of a standard module would likely move toward an extreme makeover, such as I did with an HW-101 where I replaced the analog readout with an LCD screen. But that said, my initial look at the RX110 manual may rule that out. Here is the issue: The front panel on the RX110 is sparse containing only a volume control, headphone jack, analog dial window, main tuning knob and a band switch. (See the earlier photo.)
7. Roughly speaking a 160X128 Color TFT display would fit nicely where the analog dial window is currently located. But it is the bandswitch that is the problem. The bandswitch like with other Atlas Radios is a multi-wafered/gang switched that passes through the VFO compartment as well as switching various band pass filters. Thus changing bands is accomplished with one ganged switch that controls the VFO coils as well as band pass filter circuits. So it might be difficult (at best) to simply remove the vfo hardware and in its place install the Arduino/Si5351. My initial looking (just at the manual and not the hardware) says that may be problematic. Now if we were using an external box, like a remote VFO this opens up wide the possibilities for a larger display and other functionalities. That said I will still leave the door open to an inside the radio installation. This very likely will be a two step project with the first getting to operate with an external box and based on that success consider the internal installation.
8. If a remote VFO approach is used then on the back panel of the RX110 just four RCA jacks would handle all of the chores. I already have an assembly using a 3.2 inch color touch screen already outfitted with a Tuning Knob, A/B VFO Select as well as USB/LSB Select and a TUNE Function. The four jacks would handle the LO Input, BFO Input, S Meter Output and TUNE Tone. for my conversion I would simply disconnect the VFO and BFO Intenal Connections and the power to these circuits. The net "diddling" to the RX110 would be 4 RCA Jacks.

• The output transistor used in the QRP version according to the manual is a CD2545 which of course has long ago been obsoleted. But the early Atlas 210X radios did use a pair of these in the final amplifiers for 100 watts output. The CD2545 individually is rated at 50 watts output. So a single device being run at 10 watts output means they are loafing. I have two final amplifier bricks from old Atlas 210X rigs and I replaced the CD2545 with MRF453's and they were a drop in. The MRF453 device was based on the recommendations of RF Parts who sell replacement RF transistors. BTW the MRF453 is good for 60 Watts output. It should be noted that the CD2545 is used either as the output stage in the QRP version or the driver stage in the high power version.
• For the output devices in the high power version a pair of MRF454's are used. I have seen well over 150 Watts out (below 10 M) with these devices in a RF Brick! It is obvious why the output of the CD2545 stage would be limited to 10 watts and that is so that you did not overdrive the MRF454's should that module be installed. Ahead of the CD2545 is a MPS6514 driving a 2N3866 thus there may be an opportunity to squeeze out a bit more drive and run the MRF453 (CD2545 replacement) at 30 watts output. With 30 watts of drive to my SB200, it will output over 300 watts which obviously is more than QRP and I would be heard out on the east coast on 40M. This could be exciting!

The Transmitter Block Diagram

This is where we now see how really clever the Atlas Engineers were (are) in the design of this rig. Mind you this was very likely designed about 40 years ago. My what they could do with today's technology --something I hope to do. You might want to download the TX110 manual and follow along because you can end up in the weeds pretty quickly. BTW as with the RX110 there are a lot more parts in the actual rig versus what is shown on the Block Diagram.

Strap in for here we go:

• When the receiver is used alone then an antenna plugs into a back panel RCA jack but when coupled to the TX110 module the receiver antenna port is connected to the Transmit module and this will become apparent shortly.
• Not seen on the RX110 Block Diagram is some additional hardware used only on transmit. The PC310 Board has additionally a 1N270 Balanced Modulator and a MPS6514 Transmit Gate. The Microphone Amplifier is located in the TX110 and its output is fed to the 1N270 Balanced Modualtor on the PC310 Board. The next part is very clever as the output of the Tx Gate is fed to the input of the Crystal Filter which also accepts the input from the Post Mixer Amp (2N3866). After passing through the filter and the MC1350 IF Amp stage the signal is diode steered to the back end of the Didode Ring Mixer and on to the Band Pass Filter. That is the why the receiver antenna connector is connected to the TX110. Wh2at we have is some diode steered double pass and bilateral circuitry in play. Thus the 1N914 Diode ring is both the Receive and Transmit Mixer. In my updating of the RX110/TX110, the connection point from the VFO to the Diode Ring would be bypassed and this is where we would conncet the LO Port (CLK0) of the Si5351. This is a simple replace some wiring task. The same goes for the BFO stage which is bypassed to accept CLK2 of the Si5351
• Once the signal enters the TX110 there are but three amplifiers stages MPS6514/2N3866/MRF453 for the QRP version. Internal to the TX110 are also some control circuits and the CW generation circuitry.
• One final look at the above schematic and you will note that with the actions of RLY1 and RLY2 most of the transmitter hardware is bypassed in the Receive condition save for the Low Pass Filter Networks. Thus if you have the antenna connected to the antenna port on the TX110 and have the interconnect between the RX110 and TX110 and you hear no or low signals then look to these two relays and associated Low Pass Filter wiring as being the culprit. The other issue of course is that you have two bandswitches. Thus if you have the RX110 on 20 Meters and the the TX110 bandswitch is on 80 Meters then the signals will seem low as the LPF is attenuating signals above 80 Meters. Simplicity while a virtue does have some downsides. Thinking off of the back wall --if we used an Arduino Mega 2560 (lots of Digital IO) we could have bypassed the LPF in the TX110 add an external motor driven band switch and LPF bank and thus when you changed bands on the VFO the LPF's would automatically be in place. While not a motor driven switch in my KWM-4 transceiver, when you change the VFO band the correct BPF and LPF are put in line for that band.
• This detailed analysis will be very useful as we circuit trace the signals to get both the Receiver and Transmitter operating to specifications. Note we haven't even turned on the soldering iron!