Update 12-02-2023. How to do it
Update 12-02-2023. A suggested approach for Conversion to T-50-6 cores.
[The blog post title does not involve Mary Jo @ 300 pounds and the back seat of a 57 VW Beetle.]]
A suggestion from a blog reader was to calculate the inductance of the coil and then make the calculation for the Toroid. The Coil forms in the article are Amphenol 24-5H. They are 3/4 inch in diameter and 2 inches long.
Another blog reader shared a 3 Transistor Radio Project that would work with the transmitter and can be found HERE The reader built this unit but substituted Hardened Cardboard tubes used for Model Rockets. The Cardboard tubes are sold by the Estes Company and are still in the current product line. He suggested the 0.98 inch tube. These can be found under the Accessories ~ Body Tubes!
In keeping with the plug in coil approach (likely you will not find the Amphenol parts) I think a 7 Pin miniature tube socket along with what is known as a tube saver (available on eBay) can substitute for the original. The tube saver plugs into the 7 pin tube socket and you solder the wires to the top of the tube saver . eBay Link
The reader also suggested using LT Spice for checking your substitute inductor. Good move!
There are Russian Germanium Transistors on eBay that will work in the original circuit. I think one listing has 20 transistors for about $8.Those have an Ft of 40MHz and good for about 100mw.
Update 12-01-2023. Don Stoner, W6TNS (SK) project which appeared in the August 1958 issue of Popular Electronics. See page 61. Semiconductor Space Spanner.
(Click on the link above in the heading -- have some fun reading the Ads of what was hot in 1958 in Ham Gear,)
A similar project at 5 watts circa 1967 from Doug DeMaw can be found here
You have a month -- so why not challenge yourself to building this rig and operating Straight Key Night. This radio was on 15M and Stoner even worked DX with this jewel.
Stoner is credited with being the Father of OSCAR (Orbital Satellite Carrying Amateur Radio) and also involved with the design of the Benton Harbor Lunch Boxes. He also teamed up with Pierre Goral to form SGC Electronics. Anyone own an SGC2020 QRP radio?
Yes Virginia those, are PNP Transistors.
Anyone able to convert the tuned circuits to 40 or 20 Meters using Toroid cores like the T-50-6-- send me an email?
In the interest of spectral purity and complying with your license requirements. I would follow this transmitter with a Low Pass Filter based on the W3NQN designs.
Update 11-28-2023. Another pass through the Driver Stage.
The Oracle in Florida, suggested some further evaluations of the Bias Point of the 6BA6 Driver stage.
Of note I have actually been using the KWM-1 which you can see in one of the photos. I have it tuned up on 10 Meters and have gotten some excellent signal reports. The antenna is a 40M Delta Loop.
The oracle suggested trying 3 ma of Cathode current -- well that hardly made a dent. Next I went up to 16 ma. where I had done some initial testing and got a 6 ma. swing.
The next test was at 20 Ma idling current and a re-tuning of the Driver Tank Circuit. The load is a 1k Ohm resistor connected to a 1 turn loop wound on the driver tank coil and that produces 6 Volts PTP across 1K. If I did the math correctly that is the max signal I saw or about 4.5 Milli-watts. BTW at 20 ma idling, the output shows an amazing pattern and Clean sounding output.
I may be causing harm to the 6BA6 with the bias set so high but it is a starting place. Next I will build the 6BA6 Grounded Grid Grid Final to see what would be the anticipated output.
If we saw a 10 dB gain in the GG circuit that gives us a really puny 45 Milli-watts. A 13dB gain inches us closer to 100 Milli-watts out to the antenna. 16dB gain would be closer to 200 Milli-watts.
The real nut to crack Uma Thurman (Kill Bill Movie), is the power output from the driver which still remains elusive.
Update 11-26-2023. If the post has nothing to do with radio stuff it sure gets snubbed! So Radio Stuff it is.
We had visitors for Thanksgiving so no work on the 6BA6 tube rig; but that will soon start back up in earnest next week.
So, I am suspecting my blog is much like the Pavlov Dog experiment, thus I wanted to include something about radio today. (No Radio Stuff, no Salivating!)
Behold I have a CW key -- a Johnson Speedex one of the best manual keys I own and I even have a J-38.
But the subject of this radio posting is how do you add CW features to a SSB transceiver. I have built over 50 SSB Transceivers and only two do both SSB and CW and one other is a CW only transceiver. This is not because I abhor CW; but I personally find it is harder to successfully include both modes in a rig and have it be right.
One of my SSB/CW radios is a copy of the famous W7ZOI 20M SSB/CW QRP transceiver that appeared as a two part article in the Dec 89 Jan 90 QST. Even Wes had a clunky CW system -- you held the MIC Key down to put the rig in transmit.
BUT he solved a problem that to get the proper offset he used a separate crystal oscillator that was only used for sending CW. Collins in the KWM-1 and KWM-2 had a terrible system in that they generated a 1350 Hz audio tone to key the VOX which was also fed into the Balanced Modulator.
The 1350 Hz offset is way out of kilter with the often used 700 Hz offset. But the 1350 Hz offset was the only way around a problem with spurs and the wide Mechanical filter -- 2.1 Kilohertz is relatively wide and certainly FB for SSB but for the CW Nuts, they think ten 200 Hz wide CW signals packed into a 2.1 Kilohertz pass band is awful!
I fixed the W7ZOI keying issue that when you go to CW, hitting the key caused the rig to go into transmit and there is a holding period so that the rig stays on for a timed duration.
In my KWM-4 (the other SSB/CW rig), I solved the Collins problem with a circuit out of the famous Solid State Design for the Radio Amateur (Hayward's SSDRA).
Here a separate BFO Crystal at 455 kHz used on only transmit and tweaked to give the offset. On Receive the USB BFO Crystal is switched into the path.
A NE555 timer is keyed "ON" and stays on for a timed duration (SSDRA). This powers the 455 kHz oscillator to ON, continuously during the timing period, but a follow on Buffer stage is the one actually keyed. There is also a relay that feeds the signal into the 1st 455 kHz bilateral IF (which bypasses the mechanical filter) and on up the chain like a SSB signal to the 2nd Mixer where it is mixed with 10.245 MHz crystal oscillator to produce 10.7 MHz signal which is the 1st IF.
A wide band 7 Kilohertz FM crystal Filter is in line so that the 10.7 MHz - 455 kilohertz product is not passed through the system. This also acts as a roofing filter on Receive to limit the signals passing on to the 2.1 kilohertz mechanical filter in the 2nd IF. This is another N6QW innovation.
A Si570 produces the LO for conversion to the other ham bands. You see it is complex! Then again the KWM-4 was totally my design and not a copy of another rig like the W7ZOI rig.
Vintage 1960 cheapo sideband radios simply unbalanced the Balanced Modulator and you were leap frogging across the bands on CW.
Many ardent CW ops are at crossed swords with some of the new (and expensive) high end Software Defined Radios (SDR) radios arguing about latency issues. The situation gets worse with the latency problem when the guy is in downtown LA operating CW on an SDR Radio on a remote Pacific island.
But the guys who seemed to pay a lot of attention to CW in their SSB radios was Ten Tec. As many of you are aware, I am attempting to add a Digital LO to the Ten Tec Triton II. The radio has a hokey PTO system where the PTO range is changed and either used straight on through, or doubled or tripled. A few days playing with the radio and not too bad except for 10Meters where the tuning is touchy.
My Digital LO Scheme will work as a substitute frequency generator for the Triton II, but now the CW problem.
Ten Tec uses an offset control that moves the VFO up/down on receive by simply biasing a transistor whose PN junction is like a varicap connected to the PTO. On receive this can vary the received frequency. On transmit the bias is removed and the transmitter is on a different frequency --i.e. no offset.
BUT it is not a consistent variation band to band because you are adding a fixed delta tuned capacitor (PN Junction) to an inductor that is changing value depending on the band. Hey this was a miracle in 1970 -- you know 50 years ago.
So with my external LO that can indeed be shifted to the various frequency ranges BUT how do you add the CW capability? Keep in mind in the Triton II essentially the receiver is either tuned above or below the transmit frequency and the amount is band dependent.
I am not smart enough or skilled enough to write the software where you put the main Encoder in a mode so now it is a Receiver Incremental Tuning (RIT) control and everything stays in sync.
Several RIT sketches were spotted on the Internet and were dry tested only to find they wouldn't even compile. So screws those guys!
Simplistically, what if you could write code so that you could select that the Receiver would be fixed tuned either 700 Hertz above the transmit frequency or 700 Hertz below the transmit frequency? When you transmitted then the offset disappears and you are doing CW so others can hear you. I think most of the time that on CW the USB receive option is invoked. But you could have a choice.
This in no way gives you full RIT capability but will let you do casual CW. Tune on a signal and by selection you will be transmitting either 700 Hz below or 700 Hz above the signal you are receiving.
If you are a CW aficionado, you will be terribly unhappy and disappointed -- I know I will hear: well it doesn't do this and it doesn't do that. But for someone who would like to make a few CW Contacts on Straight Key Night then it should be more than adequate.
Now another aspect to consider. Essentially the External LO would do some of the keying sequences and a reed relay would actually key the Triton II. Thus the connections to the Triton II should be +12Volts. the LO input to the TX and Rx Mixer and the Key jack on the Triton II back panel connected to the External LO.
Some magic would have to occur like placing the Triton II on a certain band and the same band punched in the Key Pad. My Triton has the audio CW filter so it would do well on CW and thus the Mode switch placed into CW narrow or wide and another selection on the Key Pad for + Offset or - Offset.
If I did it right, hit the key and the Triton goes into transmit and the display changes from the Received Frequency to the Transmit Frequency.
Now I haven't actually written the sketch, but my Klingon Mind Meld Mode says something like this should work.
When you return to SSB then the offset crap is stored away and you transmit and receive on the same frequency like usual.
Now Ten Tec must have realized their offset system might be marginal as the manual says one option you have is to defeat the Offset by plugging in 9VDC on some board.
Have you salivated enough for today?
Update 11-25-2023. How to get work done properly and inexpensively in trade for Food!
Hood support Struts 2006 Avalon
We also do Ham Radio Stuff Click here!
I drive an 0ld car for several reasons. The 1st is that it still is serviceable and the second given the cost to have the XYL in Board and Care is very expensive and new car doesn't fit in the budget. The car in question is a 2006 Toyota Avalon.
There are times when it is necessary to transport the XYL to medical appointments (next Tuesday) and so the family car needs to work. The issue of old: Do I fix it or buy new?
About 6 months ago the Trunk Lid Torsion bar assembly failed and not easy to open the trunk. The XYL uses a transport chair and of course stowed in the trunk. It can fold down and put inside the car in the back seat. Because the XYL is on oxygen she also rides in the back seat. Shades of Driving Miss Daisy!
But for longer trips like to a medical facility in LA, I have a care giver ride along as it is hard to tend to her needs (back seat) when going warp speed down the 101 Freeway. Thus the chair has to go in the trunk. That was a $1400 fix.
While that was being serviced I was "advised" that the hood support struts had failed and that somehow the cap cover for the windshield washer fluid reservoir was missing. That would be an additional $300. I said NO.
That is when I hit upon an ingenious idea after watching a TV Ad for eBay motors. My #3 Mechanical Engineer, Car Nut son would be visiting for Turkey Day and so I buy the parts and he installs them in exchange for food. The reservoir cap looked pretty easy but the struts looked a bit more involved. So I trade him lots of food for some car work.
The struts cost $20 and the cap $10. The total elapsed time was 15 minutes including a short break while he took a phone call. So $30 and 12 minutes was 1/10 the cost at the dealer! Oh, some service gorilla got the hood release cable disconnected from the dash mounted release latch -- 10 seconds to fix.
Needless to say I am really pissed! I do not mind paying for services rendered. But that is robbery! In the past 17 years I calculate that the hood has been opened maybe 100 times. I have the car serviced 4 times a year and a few times I looked under the hood. The struts fail because of the seal that keep a fluid that is compressed under pressure.
While he was here he also "tuned up" the CNC machine (more food to take with him). The tune up included some new and bigger X, Y drive motors and he also showed me some tricks to change the cutting program essentially on the fly. We even made some parts for an upcoming project.
The $250K I spent sending him to college is really paying off. Food for work -- it works!
Update 11-23-2023. Happy Turkey Day!
I personally have much to be thankful for this year despite many bumps in the road. I do hope blog readers today are all enjoying a great day. Hopefully with lots of family around.
But when I think back on past Thanksgivings, I smile because of this day. At one point in my life I was an outside sales engineer for a distributor of automated industrial controls. A customer contacted me and needed some help.
This customer bred Turkeys and that was a problem! Turkeys unlike other animals do not have a predisposition for mating. So no mating no little turkeys. The project premise -- speed up Turkey production.
This customer was working on a process to artificially inseminate turkeys and it involved lots of controls. Things like a conveyor belt, timing and process control mechanisms and pneumatic guns to do the insemination's. It was a salesman's dream come true!
Don't know if it worked but I sure had a fat commission check for Christmas!
Update 11-22-2023. What Chat GPT has to say about GG
ChatGPT can make mistakes. Consider checking important information.
Update 11-21-2023. The 6BA6 Final in a Grounded Grid Configuration
Based on the input from the Oracle of Florida, I sketched out what a Grounded Grid 6BA6 schematic may look like.
Thinking ahead to the Final Amp Stage, while I make the several adjustments to the Driver Stage.
The Pi Tank Network is from a junker SBE-33
Stay tuned: The Goal is a 44DD, two watt output from the Final.
Update 11-20-2023. The Oracle has Spoken. More on the moving the 6BA6 from a 29A to a 44DD.
Last week I was really excited to see the Cathode Meter move in unison with the Voice Peaks and then realized it was a rather puny response. But even at a 6ma swing (22ma peak) and 260 Volts DC on the plate I was reminded that the plate dissipation was (0.022 x 260)/2 or nearly 3 watts. I was on the ragged edge.
The Thermatron Oracle made some suggestions which I will undertake.
- The Bias is set improperly as I should look at setting the resting current to maybe 3 ma. If you can swing from 3 to 22 ma (about 19ma) and at 260 Volts DC on the plate that is some serious output.
- A look at moving the stage to a Triode configuration (The 813 is often configured like a Triode to handle more power ). Maybe even consider a Grounded Grid arrangement.
- The matching into and out from the stage. Now The output from the SSB exciter is at 50 Ohms and I should look to using a Broad Band match to mate with the higher input grid impedance of the 6BA6. Grounded Grid would be closer to matching the Ins/Outs.
So the Oracle has spoken and now we need to prepare a plan. But in the meantime I can think about building the Triton II LO.
Update 11-19-2023. I am seeking some professional help on how to get more juice out of the driver stage!
It is like when in High School your girl friend was a set of 29A's and no where near Melanie, the best looking cheerleader. There it is sitting there with a Cathode current swing of only 6 Ma. Some thoughts include better matching to the input side and a better output matching.
Maximum power transfer theory demands matched loads. On the Solid State Side we all know about impedance transformations and for the vacuum tubes that has to follow.
I am awaiting a return email from the Oracle of Florida with perhaps some suggestions on how to swing perhaps 15 to 20 Ma from the driver versus the (29A) 6 ma.
In parallel I just acquired a Ten Tec Triton II. I had one about 2 years ago which I passed on to a ham just starting out in the hobby. He in turn for his next rig bought a FLEX 6600 --he went from free to $6K. He still has the Triton II as a backup rig.
From N6QW ~ Triton II Digital LO
I like the Ten Tec radios as the company's roots (modules) are evident in their products -- you can isolate circuit boards which make trouble shooting and modifications a whole lot easier.
The Triton I and II are particularly interesting as they use a PTO which is very stable and then this PTO is not mixed with any crystals to shift the range but instead inductor networks are switched in parallel with the PTO to shift its range using a simple Colpitts oscillator. From there the output is either straight through, doubled or tripled to inject the signal into the RxTx Mixer board.
Getting the range to be proper for each tuned network for each band is much like an evening with Mary Jo (300 pounds of fun in the back seat of a 57 VW Beetle).
Above is the VFO Schematic with L11 being the main PTO. On 10 Meters L9 and L10 are switched in parallel with L11.
The mating dance is with the combo of L9 and L10 that one coil sets the lower band limit and the other sets the band spread. The trick is just the right combo so that you come close to starting off at the band edge and hit all of the 100 kHz points along the dial string. This is best described as getting Melanie (The prime grade Cheerleader) in the back seat of the 57 VW Beetle.
On 80M the VFO output midpoint is 12.750 MHz followed by 16.250 MHz for 40M and 5.250 MHz for 20M then 12.250 MHz for 15M and 20 MHz for 10M.
This is a situation ripe for adding an Arduino and Si5351. A bit of a trick in the programming as you display the info on an LCD. On the display for 80M and 40M the VFO subtracts the BFO to give LSB in a BPF and for 20-10M the VFO adds the BFO in a BPF.
Ten Tec and others call this Normal as the subtraction in a Band Pass Filter produces LSB and the addition in a Band Pass Filter produces USB.
Of course I have ginned up a suitable outboard VFO which with a Pro Mini and Si5351 does all of the heavy lifting. Stay tuned for a video while I await to hear from the Oracle from Florida.
Update 11-16-2023 Revised. We did go back to known stuff.
We made some adjustments to the Driver Stage by reducing the Screen Voltage to 150 Volts DC and we added a Cathode Meter so now we can watch the Cathode current swing. We can see a 6-ma swing and for the final we have to have an 18 Milli-amp swing for 2 watts. So, we are not close but certainly an improvement from yesterday. At 15ma Cathode Current the Grid Bias was set at - 6VDC. That does not match the characteristic curve load line so yet another item to find out the why.
At 100 Volts DC on the screen and at certain Grid Bias Levels we actually saw downward modulation. Not Good!
There was a definite output pattern improvement but not what I would call a real Amplification -- the drive level is more than the output!
Nice to see the meter swing!
Update 11-16-2023. Go Back to the known stuff.
I had postulated some courses of actions, and the 1st is to look at Class A Operation. The second is to look at the Characteristic curves.
My Screen voltage is 200 Volts, and the Plate Voltage is 260. For a Class A Amplifier, the Plate Voltage is OK, but the Screen Voltage is way too high. I also added -3 Volts of Bias to the Grid circuit and that improved the output pattern tremendously so that it follows the input! But it is not amplifying like it should.
Today I will modify the screen voltage to 100 and then video the results. For a load line draw a line from 11 Ma of Plate Current down to 260 Volts DC and then look at the crossing point at -3 Volts of Grid Bias. That should put you in the middle of the operating range.
You can see why transistors may be a bit easier to circuit develop.
Update 11-15-2023. Not so good a test! The Driver Stage is not operating properly!
A Quick and Dirty Video of a Dummy Load...
A close up of the Driver Stage
The Low Level SSB Transceiver
The Driver/Final Board
I tested the Driver stage with the low level output from the SSB Transceiver. The results were not good!
The tube was not amplifying as such but I did get an output from the link coupling terminated in a 50 Ohm resistor.
As such the output waveform does not look linear or clean and was like more of a dummy load than an amplifier stage. I did increase the level of the input signal and the output level increased but it was not amplifying as such.
So that now goes to the WHY?. The plate voltage is 260VDC and the screen voltage is 185VDC. Typically the 6BA6 runs with 68 Ohm cathode resistor but I used a 51 Ohm in hopes of having it run a bit hotter. The grid resistor is 270K.
Some things to evaluate:
The operating point may be as such that it is in a region not best suited for amplification of SSB signals.
2. I am thinking that I may need to run the stage as a Class A amplifier to improve the linearity.
3. I may be peaking the tank circuit but it may be on a harmonic and not at frequency. I just took a coil that looked about right and tack soldered a 25 PF trimmer cap across the coil. That needs looking at --it does peak up but not a huge peak.
4. More testing to follow. This is why "fun" is the root word of dysFUNctional!
5. Lastly I tested all of the tubes but in fact this tube itself may be an issue.
6. I did verify the Pin voltages so if you get an itchy typing finger that was checked--and Pin 5 is the Plate and Pin 6 the Screen.
Update 11-12-2023. The Safety Stuff.
When I built the wireless transceiver I used a Heathkit HP-23 supply for the power and along with that an external PC board including some power dropping resistors and adding a 150 Volts DC regulated source to the rail. Now we were cooking. However I always was concerned because the PC Board was essentially exposed and that this presented a safety hazard.
In an earlier post I mentioned I would tackle that safety problem and I did by building a metal shield over the PC Board which is then screwed down to a wooden board.
This post details how I built that shield. I bought a 10 piece bag of really absolutely thin (useless is more like it) aluminum sheets 10X4 inches.
However by using a process called "hemming" you double over an edge which makes it really strong. The secret goes back to 9th grade where our shop teacher taught us some great things about sheet metal fabrication. The real trick lies in the layout and visualizing what gets bent first.
I bought a bench top metal break from Harbor Freight about 15 years ago and it was less than $30 -- I have certainly got my money's worth.
So look at the photos and see how I did it.
It is also a good idea to drill any mounting holes at this time --including using a 1.25 inch hole punch. (A Greenlee Chassis Punch for tube sockets.)
This Big Hole is to pass the cable through for the power switch to the Heathkit HP-23
Now we do some initial bending to form lips that will be bent over in the hemming process.
Use a Rubber Mallet to bend over the metal.
The later photos show the Pi Network and the Cathode meter install. The power switch controls the On/Off of the HP-23.
When You Know Stuff You Can Do Stuff!
Next will be a two stage test after connecting to power leads to the 6BA6 board. The 1st test will be to just evaluate the 6BA6 Driver Tube and then the follow on piece will to have the Final and Driver in the circuit.
PS I made no attempt to make this a perfect bending operation and total elapsed time was about 40 Minutes. In retrospect the thin metal may have not been a total mistake as the "U" shaped cover for the Shirtpocket transceiver used this same stock.
I just saw a photo of a ham who built "The Simple SSB Transceiver" as modified by the Vienna Wireless Society. This ham gets extra bonus points as he homebrewed his case using this same sheet metal approach. Now there is a total homebrew radio!
Update 11-11-2023. Moving to installing the final amp stage.
This was a fit check and now to anchor the assemblies to the wood base and do the final wiring.
Some stuff takes time and then it all falls into place.
Update 11-10-2023. The Driver stage is built and awaiting testing. Happy Birthday US Marine Corps!
Caution: We now will be working with voltages in excess of 12 volts. Good practice dictates to physically isolate any exposed high voltage wiring and one hand behind your back. As I complete the Driver / Final board you will see how I approach high voltage safety.
Spent today building the driver portion of the RF Linear Amp Stages. I tweaked the SS xcvr so that out of the BFR106 I am now getting 700 Milli-Volts Peak to Peak which is just about 1.2 Milli-Watts. (0.7^2 * 2.5).
I am hopeful of getting about 5 to 6volts PTP out of the driver which would range from 60 to 90 Milli-Watts output. If we can do 15dB in the final stage that would give us around 2 watts to the antenna. A bit of math: 2000/62.5 = 32. 10*log(32) = 15 dB. If we can get 90 Milli-watts then at 15 dB gain the output would be 2.88 watts.
How I got there: since the log of 32 X 10 = 15 dB then we know that the power output divided by 90 has to equal 32 -- so a multiplication of 90X32 = 2880 Milli-watts. The plate dissipation of a 6BA6 = 3.4 watts.
The next task is to test the Driver Stage for a stable output and any tweaks that would get me closer to the 90 Milli-watts out of the stage.
The final tank circuit will be mounted on a small vertical panel and a panel has been built for the Cathode Meter.
A preview of things to come includes adding a TR relay and actually operating the 6BA6 stage with the SS transceiver. The next pivot would be to install the already built 6BA6 audio stage in place of the LM380 Audio amp circuit. I would also look to testing other pieces like the 6BA6 Product Detector in place of the ADE-1.
The simple concept is to take known working circuit modules and to utilize these to test the new and unproven modules. Heathkit was very successful in this approach.
Update 11-09-2023. Getting the Low Level Drivers Stage peaked up.
We know this part is working FB and so now can work on the Driver and Final stages.