New Technology for 2020 ~ Life after Field Day

Moving forward by 64 Years from 1956

July 6, 2020 ~ An SDR in your Garage!

A regular reader to this blog shared that he is developing a greater use of a Raspberry Pi to assume more of the workload in a configuration similar to my RADIG.

In the RADIG the RPi3 does the Software part of the SDR. But everything else remains as "analog blocks".

One of those blocks is the USB Synthesizer/controller from SDR-Kits LLC in the UK. Essentially this block is a small board with a ATTiny 45 (or 85) microcontroller and a Si570 along with some voltage regulators and other small components. This board connects to the RPi3 via USB and responds to I2C commands so that at least three important functions are accomplished (there may be more).

  • First is changing the frequency of the Si570 whose output is at 4X. Downstream is a 74AC74 on the main RADIG Board that creates two quadrature LO signals at the operating frequency (read divide by 4).
  • Secondly commands from Quisk like PTT are translated via a switching transistor so that control of TR is done in other parts of the transceiver
  • Finally you can input to this board in the form of CW keying
The microcontroller is responding to I2C signals from the RPi3 -- so what if you simply transferred that function to the RPi3? The 40 Pin GPIO header has an I2C output which if you switched that workload could output the signals directly to a Si-570. But why Si570 -- why not control the Si5351 from the I2C. If my memory serves me correctly the control of the Si5351 via a RPi3 was done several years ago by a UK ham and it was over the Internet! One of the GPIO pins could serve as the "TR" control pin. Other pins could be the CW keying input.

Fortunately the Qusik software has variants to accommodate various rigs. I call my rigs a "Soft Rock" but one of the subset variants of the Quisk is to have remote control. So how about your RPi3 controlled rig at your home QTH and your are at Starbucks with your new mask fitted with a straw port and you are socially distanced but working FT-8 from your iPhone using your rig at home. 

Using the RPi3 to eliminate the USB Synthesizer would be a huge step forward. Better yet would be to use the 8 GB RPi4 as this should easily handle the added workload. Next would be the Sound card box. There is a sound card sold by FePi that plugs into the 40 Pin header. It has the all important Line In, Line Out ports. We would not want to plug this into the GPIO header but we could make an adapter to plug into selective pins. It has a much smaller foot print and frees up a USB port on the RPi. 

So what is happening here is that the RPi3 (4) is assuming more of the workload and the foot print and number of external boxes are shrinking. Someone who has the patience to design a proper circuit board would make things even more compact. My discovery of using the Mean Well SD 25A-5 power supply would now make a single 12 VDC source supply much easier on the power management. 

Most of the remaining circuitry would remain fixed; but importantly as you change bands in the Quisk software three GPIO pins could be decoded to provide automatic band switching for 8 banks of BPF's and LPF's. Change bands on the Quisk and the right set of filters are in line. 

Regrettably I do not know how to do all of this, especially the RPi software changes; but I suspect among the readers of this blog pieces are already in place. 

Now the Bonus -- if you get this working on a RPi3 (4) the ASUS Tinker Board is literally a drop in. The ASUS TB is more powerful than an RPi3!

Pete N6QW

July 5, 2020 ~ Time to get serious!

We are into the second half of 2020 and this will be a very interesting time as we continue to struggle with Covid19. We must look to science for the answers. Wear a mask, employ social distancing and use your head. For those living in Amsterdam --well you know --no heavy breathing or kissing!

But this is also the time to be thinking about lining up a Fall/Winter project. Building an SDR Transceiver using a Raspberry Pi3 is within your reach and the results are astounding. Parts have come way down -- for less than $10 you can have two ADE-1's and the Si5351 can be found for around $5. The sound card is $15 and the USB Synthesizer is around $30. A quick add up and that is around $150. So start saving now and think about having a homebrew SDR in your shack by Christmas!

Pete N6QW

July 4, 2020 ~ Happy 4th of July!

Independence Day here in the USA and I guess in the UK too as the 30,000 pubs across Britain were reopened today. I sure hope everyone practices social distancing and are wearing masks. 

I also read on the Microsoft news feed where the sex workers in Amsterdam are back at work where the restrictions include no heavy breathing and no kissing. Wow, their contact tracing will be most interesting!

Because of my work on the KWM-1 which is on a short hold while I await some replacement electrolytic caps, I have once again drifted back to 20 Meters. My beam is broken but the 40M Delta Loop sure likes to load up nicely on 20M. 

This in turn prompted me to update the 2nd ZL2CTM rig and move it from 40 Meters to 20 Meters. Only three changes were needed and includes swapping out the Band Pass Filter, Low Pass Filter and changing the Boot Up frequency of the Teensy 3.5. My Teensy has code where essentially the Si5351 (itself) will provide signals from 8 kHz to 160 MHz but that is one continuous LO. Since this is a single band rig -- easier to boot up in the chosen band of operation.

Long ago using LT Spice,  I came up with a Band Pass Filter design that by changing two 150 PF (COG) caps and retuning five trimmer caps the BPF design could work on either 40M or 20M. So the BPF change involved a simple unsoldering the ground ends of the two 150 PF caps and a trimmer retuning. 

For the LPF design I used the stock W3NQN 20M LPF filter design that employs two each 220 PF, and 330 PF capacitors along with a single 47 PF and three T-30-6 cores. I built the 20M LPF on a board the same size as the 40M LPF and with mounting holes at the exact same location. Read a drop in!

Thus if I get tired of 20M and desire to go back to 40M --simply solder back the two 150 PF caps and retune the five trimmer caps. The swap of the LPF requires the unsoldering of two wires and removing two sets of nuts and bolts and replacing the LPF board --two solder connections and replacing two nuts/bolts. A quick reprogram of the Teensy and back on 40 Meters.

Now everything thing has a price and I note the sensitivity on 20M is down a bit from 40 Meters as the Receiver RF amp stage is broad band and tends to favor 40 Meters. So going higher than 20 Meters say to 17 Meters likely would need a refined LT Spice simulation of that this Amp stage as it also does double duty through relay signal steering as the Transmitter Pre-Driver. Thus the transmit power output would be affected from less drive and the IRF510 itself tends to drop off power as you go higher in frequency.

In anticipation of a possible change from the IRF510 to the RD06HHF1 RF FET, my linear amp board already has a LED fitted to the 78L05 regulator so I can develop a higher bias voltage required by the RF FET. It is a simple installing the RD06HHF1 into the IRF510 connections and you have the change over. It will also require a readjustment of the bias (upward).  BTW when you transmit it is so cool to see the Red LED light up. Now if I could only find the cool Juliano Blue LED that is "somewhere" in the bins that would be the ne plus ultra!

There are the 13 colony stations operating during this several day period and I will attempt to work one or two; but of course I am not a contester or collect paper. W9IMS (Indiana Motor Speedway) is also handing out certificates this weekend too.

Stay at home this Independence Day so you can be here to see Labor Day! Wear a mask and practice social distancing! No heavy breathing and no kissing for some activities at certain locales are also  restrictions. Happy 4th of July!

Pete N6QW

July 3, 2010 ~ Smart Collins Engineers

Let's face it the Collins KWM-1 is eligible for Social Security based on its age! So like Rod Stewart's  "Maggie May" the KWM-1 has been around the pole (more than a few times); but something is truly amazing about that rig and its design. 

With today's modern radios (ICOM, FLEX Elecraft. Yaesu, Kenwood) few are ever opened up to view the innards. There are good reasons not to open these radio appliance boxes. Firstly, many of the owner's simply lack any technical skills other than pushing menu buttons and secondly with all of the SMD components hard to see what is on the board(s). Thirdly I guess is that often specialized test equipment is needed to service these jewel's. Fourthly, the repair business is very lucrative, so by design you are forced to send the radio back to the manufacturer. 

I have heard tales that a trip back to the manufacturer for servicing (even for simple problems) could result in cost as much as 50% of the purchase price. Hey at a shop rate of $100/hour --five hours is 1/2 of the cost of a  ICOM 7300. Any time you open a radio -- two hours is minimum as they not only service the problem; BUT then they check and realign everything even if it is not needed!

Now the Collins of 64 years ago was backed by some really smart and talented engineers working on the products. They literally thought of everything including  a manual that had servicing sections that included voltage and resistance charts so that the owner could in fact perform servicing. Also evident -- replacing a bad tube was a fairly simple "jerk, pull, replace". True some tubes required further action as was the case with the finals where a replacement likely would require neutralization. 

In my KWM-1 manual there is strange note about replacing finals that essentially says if upon replacing the finals if there is no apparent unwanted oscillation Don't Screw With The Neutralization. Have you wondered how you would neutralize an ICOM 7300. See, you owner's of a 7300 probably haven't realized -- no need because of the more modern designs. 

But there was a  lot of packaging engineering that went into a Collins Transceiver as it is jam packed with parts and a great deal of "knack" is needed to remove parts and assemblies. That is some of the difficulty with my attempt to remove the PTO. There is an easy way or a brute force way. Obviously the brute force may result in broken parts that simply are unavailable. So that is a concern for on how to service the PTO without breaking something. I now have enough info after polling the greater Collins community on how to do it in manner that protects the existing PTO. First rule -- ask first!

In a recent Soldersmoke Podcast, Bill, N2CQR mentioned about a Collins owner who only liked to work on the separate units (75S3 or 32S3) as there was reasonable room to service the units whereas the KWM-2 he stated was a "Jumble and Jungle" of parts. 

The KWM-1 is well described as a Jungle and Jumble of parts. I have long held the thought that Collins employed very dainty ladies with exceptionally small hands and thin fingers to assemble the radios. No Gorilla Guys on those assembly lines. 

That said how many computers do you have in your home/shop/shack that are more than 25 years old (that is 1995)? Collins Radios endure because they were designed to be so.

Happy 4th of July -- If you want to see Labor Day, stay at home on Independence Day! 

Pete N6QW

July 2, 2020 ~ Operating the KWM-1

So I have had a chance to check out things and  found there is an issue with the PTO. Problem #1 is how to extract the PTO out of the radio. Have gotten some good input from various Collins experts but certainly don't want to use a sledge hammer when a dab of liquid wrench would do. Still working that problem. If I can get it out, I am certain I can fix it!

I have hooked up an external oscillator and that was done by taking a 7 pin tube with good filaments, but low emission, wherein I cut off all of the pins less the filament pins. Pin #5 is normally the plate and I just stuck one end of a 10NF 500 VDC ceramic cap in Pin #5 of the socket and then connected my Feel Tech sig gen to the cap to the other end. 

The PTO range is 3.455 to 3.555 MHz --only with the Feel Tech it seems to tune backwards so that is good to know when I fix the 70K-1. Thus 3.455 is the high end of the range. That also tells me since the PTO seems stuck near that frequency then the core is not moving out of  the sleeve. This was a really critical experiment.

I also checked my stock of crystals and found some 11.04 MHz crystals. Using the magic decoder ring 2X11.04 = 22.08 now subtracting 4 = 18.08 MHz which is at the low end of 17M and tuning up (when the PTO is fixed) means it would go to 18.180. BOOM 17M FT-8 and the phone band on the KWM-1

Have not checked out the transmitter --small steps grasshopper!


Pete N6QW

June 30, 2020 ~ Spotted on Some Road Signs

I wish some people (like the emperor) were fluent in silence!

Latte (n.) Italian for you paid way too much for a cup of coffee!

Here we are at the end of another month with no end of the pandemic in sight. Yet  a large number of Americans are hopeful in 126 days may signal the end of one of the greatest problems this nation has ever faced. 

June 29th, 2020

Today we are supercharged with the latest radios and their feature offerings. Typically you hear that the rig on the other end is an ICOM 7300. Weighing in at around $1000, it is  super star, employing the latest SDR technology and an undeniable track record of high quality and superlative performance. Notably I am also starting to hear a lot of the lower-end SDR radios from China on the bands. Costing less than 1/2 the price of the ICOM, their impact is soon to be seen and felt! (Hey ICOM there are some new kids on the block!)

But some 64 years ago, there were radios that supercharged the ham community which also cost $1000; but in today's dollars it is more like $10,000. Chief among those was this jewel.

This is the Collins KWM-1 which arrived on the scene about 1956. The historical folklore has it this radio was independently developed entirely by one Collins engineer, Gene Senti,  in his basement no less. The engineer (actually I recently found out was in the management chain) it seems bought a NIB 75A4 with his employee discount and modified it from a receiver to now become a transceiver. A visit to his home by Art Collins and the rest is history. Imagine what went through Art Collins' mind that day -- literally free development costs and a virtual gold mine for sales.

This was a limited production run of I think some 1200 to 1300 radios and was replaced by the Collins KWM-2 about 1959.

Let's look at some specifics about the KWM-1 with the first being it was a transceiver! Up until this time the typical SSB ham station was two large boxes weighing hundreds of pounds (like the 75A4 and the KWS-1 known as the Gold Dust Twins). 

Now in a small box was the capability to communicate world wide. One of the most famous KWM-1 radios was installed in Gary Power's, U2 Spy Aircraft that was shot down over Russia. 

The operating range was 14 to 30 MHz, USB only and while it had a PTO, the KWM-1  operated in 100 kHz slices in that 16 MHz range. For instance on 20 Meters it would operate on the stock frequencies of 14 to 14.1 MHz, 14.2 to 14.3 MHz and 14.9 to 15 mHz. The tuning range is set by a front panel Crystal pack. 

On my two KWM-2's the tuning range is 200 kHz per slice and for 20 Meters is 14.0 to 14.2 and 14.2 to 14.4 and 14.8 to 15 mHz. I guess the 15 MHz is so you could listen to WWV. 

Now special crystal frequencies could be ordered for the KWM-1 to give other ranges such as 14.25 to 14.35 MHz, which would catch some of higher end of 20 Meters. 

When I was in the US Navy Seabees, all West Coast Mobile Construction Battalions had a Collins KWM-2 and 30L-1 for phone patching back to Port Hueneme, CA. The phone patch operating frequency was 14.335 MHz, so indeed coverage in that upper range would be needed. Many KWM-2's have been changed ( a simple crystal change) so that the 2nd range is 14.150 to 14.350 MHz which aligns with today's phone bands on 20M.

Here is the innards of the KWM-1 Crystal Pack. There are ten slices of coverage including three each on 20 and 15 Meters and 4 slices on 10 Meters.

Forthwith is the magic decoder ring on the Crystal Frequencies. One of the crystals is marked 9100 KC. Now if you take that frequency and multiply it times 2 (X2) you get 18.2 MHz and next you subtract 4 and the result is 14.2 MHz -- the lower end of the 100 kHz slice. Another frequency is 9125  KC -- doing the math give 18.250 MHz and subtracting 4 = 14.250 Megahertz. So this is a non-stock crystal; but gives you the 14.250 to 14.350 MHz spectrum. 

Using the MDR ( Magic Decoder Ring), if you could find a 11.046 MHz crystal this would give you a slice from 18.092 to 18.192 Megahertz. Boom 17Meter FT-8 and SSB operation. Oh such a crystal is available and costs 60 cents at Mouser. It is an HC-49 so you would have to retrofit it inside a HC-6/U crystal can which is easily done. 

A crystal at 16.250 MHz gives operation at 28.5 to 28.6 Megahertz. You would probably want a 16.15 Megahertz rock so you could operate 28.3 to 28.4 Megahertz. 

If you look at the accessories for the KWM-1 the standard Crystal Pack (for transceive operation) can be replaced by what is called the DX Adapter which enabled transmitting on one of several frequencies, while receiving on just two frequencies (in the 20M band). Essentially you now have a channelized crystal controlled radio. This really seemed strange as you actually  have to physically modify the  KWM-1 to use this adapter. 

A former Collins employee recently suggested to me that this modification was for the U2 application. Flying the U2 you simply didn't have time to spin the dial. Channelized operation controlled from the cockpit was mandatory. It seems like the KWM-1 was mounted behind the U2 ejection seat. 

Imagine an ICOM 7300 mounted in a U2?

In one of the older vintage late 1950's ARRL Mobile manuals there is an article that shows how it would be possible to modify the KWM-1 to operate on 75 Meters. Thus in short order the KWM-1 evolved into the KWM-2 to give all band operation and by using non-stock crystals literally operate from 3.5 to 30 MHz. There are at least one or two ranges that are not recommended for the KWM-2 -- I think somewhere near the 60M band is one such area. One of my KWM-2's is now on 17 Meters. Adding LSB or USB plus all frequency operation gave it greater market appeal.

On the hardware side there were many "repeat" tubes like three 6BA7's and three 6AL5's which are dual diodes. The choice of the 6AL5 tube I think had to do with filament current leveling. The KWM-1 filaments could be operated from 6.3 VAC (standard on the 516F1 supply) or 12 VDC (DC supply) or 28VDC (DC Supply). Depending what pins were used on the power plug to feed the filaments enabled the three choices of filament voltage. I suspect the 28 VDC supply was to enable operation in the U2. 

A side note here. If you have a Heathkit HP23 Supply and fitted (configured) the power plug like you would with the 12 VDC supply, the KWM-1 or the KWM-2 could be operated from the HP-23. The filament draw with 6.3 VAC is 10 amps and with the 12 V supply only 5 amps. Interestingly enough, two pins are used to deliver the 6.3 VAC filaments as the wire size that fits the special power plug that would carry the 10 amps is too big for the pin. By using a size 22 wire (on two pins) each can carry up to 7 amps. So plenty of safety margin and using a smaller wire size.

My first close up look at a KWM-1 was in 1963 when I was stationed on Midway Island. The mission of Midway was to support the Airborne Early Warning squadrons who flew out of Midway. These crews flying the Convair Willy Victors rotated to Midway on a two week basis and one of the pilots on those crews regularly carried a KWM-1 to Midway and that is where I saw it in operation from the BOQ (Bachelor Officer's Quarters). I was stunned as my rig back in Pennsylvania at that time was a clunky DX-100 and a SX-99 -- on AM no less. Here he was working DX with a KWM-1 using a chunk of wire thrown out the window to an Australian Pine tree. Wow.

I have often thought back to those days on Midway and the KWM-1. It was a thrill to see it in operation and given the limited production run not typically available on the market today. Thus the cost of a KWM-1 is unlike  the KWM-2's or KWM-2A's which now can be found from $400 to $600. 

A very nice KWM-1 commands a King's ransom. On eBay currently is a hunk of junk with virtually no tubes, missing the Mechanical filter and has the DX adapter installed. It has been listed for the last six months for $900. I have been tempted to offer $100 --- just might do that.

The photo you see above is  now  in my shack and is S/N 673, so about mid-way through production. It will need some work; but I did not pay $900 but more in line with what you can buy a KWM-2 today. 

There are several tasks I must undertake to get this puppy perking and barking! My usual process is to start with a full visual inspection. One item missing is the perforated metal cover over the final amplifier compartment. I know how to fabricate a replacement so that is not an issue. 

Next I will pull all of the tubes and test each of them adding a bit of deoxit to all of the tube pins.   Luckily I have many replacement tubes in my tube stock so we should be good there.

Next I want to look at the resistance charts for all of the measuring points. There undoubtedly will be out of tolerance parts. Again I have a fairly decent stock of 1/2 watt resistors and HV capacitors. Luckily the deal also included an original manual where you can actually read the print.

In parallel I am building an adapter power cable so that I can use my KWM-2 power supply to power the KWM-1 (the 516F1 and 516F2 are almost identical). I will use a Variac to bring up the power slowly to reform any capacitors. The initial testing will NOT include any electrical connection to Pin 2 (the 800  VDC HV pin). 

My adapter cable is about two feet long and inserted btween the KWM-1 and the 516F2 cable and has been designed so that I could power the KWM-1 from either the 516F2 OR I also have a heathkit HP23 which has been frequently used as a power source for the Collins gear. All of the wiring to accommodate either supply is outside of the adapter minibox and dependent to the cable coming from either the 516F2 or the HP23

There is a mechancal problem with the 70K-1 PTO. I know how to fix it. What I don't know is how to remove the PTO --without breaking something in the process.

Stay tuned as I relive my days on Midway Island.

Pete N6QW

Popular posts from this blog

21st Century Homebrew SDR SSB Transceiver Project

New Technology for 2020 ~ The Phasing Transmitter

New Technology for 2020 ~ It ain't over until the Fat Lady Sings!