All Things Wireless & Letterpress

All Things Wireless & Letterpress

Thursday, April 30, 2015

Update: The "Mini Bandmaster"

I suppose the first thing I should write about is the blue transformer.  Many of you who build audio equipment will recognise it immediately as a power transformer from Edcor.  And you would be spot on.  Two months ago I ordered a 360-0-360v / 440mA power transformer from them to serve as part of a larger power supply for the Mini Bandmaster.  The reason was that I was having a very difficult time getting more than five watts from it.  I was using a much lower current transformer, which I initially figured was the problem.  And to be sure, it was part of the problem, but only a part.  I will explain.

I built this "MOPA" (Master Oscillator Power Amplifier) type xtal controlled transmitter many times over the years, without a hint of problem.  Always with surplus or "junque box" parts.  This project was no different.  The original circuit comes from the 1954 ARRL handbook.

One of the unique features of this circuit was the oscillator output to the 6V6 final.  The author of the original article has the builder utilising a 100uH RF choke to serve both as a Radio Frequency choke, and also to tune the oscillator output to 5 MHz. (hereafter referred to as "Mc".  Hertz rents cars!).   This was to tune the oscillator output to 5mc, which is halfway between 3.7mc and 7.1mc, the then current Novice frequencies.  He states that this provides sufficient input to the final to drive it properly. It is a compromise. This eliminates the need to tune the grid input on the final, reducing the parts count, cost, and complexity of construction and operation for the beginner.  And to be sure, it did, and it does. When it works.

The only problem was that in this case, it didn't.  I purchased two 100uH chokes from my favourite Military Surplus shop down in Orlando (Skycraft Military Surplus), installing one, as per usual.  When the construction was finished, the power output was very low, only five watts.  This rig should do at least fifteen watts.  

I removed the 6V6G, and inserted a 6L6G, only to find the results being the same.  Five, maybe six watts.  The best dip I could get was 60mA on the 6L6G, and 49mA on the 6V6G.  

I rebuilt the pi-output, thinking this was the culprit.  No change.  I concluded that the power supply just did not have the moxy, so I set about rebuilding it with a transformer which I knew could handle whatever was needed.

Here it is, in mid construction.  I figured that I could use this for future projects handling 807s and 6146s, so it was a great investment.  I built the chassis from angle aluminium, which provides a very, very sturdy foundation to build upon, and allowed a space in the center to both mount an octal socket, and feed the transformer wiring through.  It came out nicer than I had allowed myself to hope.

But when it was finished, I realised only a scant increase using either the 6V6 or the 6L6!  This had me going back over the schematic.  What did I forget?  What was that one thing missing?  I went over all my bypasses, all my tolerances, measured all voltages, everything was there and functioning.  WTH??  (What The Heck?....No, the F-bomb will not work it's way into my blog!)

Bry, AF4K, wondered about that 100uH RFC.  I told him I used these all the time, no problem.  He said that he always used a standard 2.5mH RFC, and also had no problems.  So, I pulled it, and stuck a 2.5mH RFC in there.

Wonder of wonders.  10 watts on the 6V6G!!  That 100uH choke was the problem.

The only thing I can think is that these 100uH that I bought were used for something other than RF.  Maybe audio.  I don't know.  But one thing sure: it was the problem!

So, as long as I had the rig open, I thought why not just go all the way and actually place a tuned circuit in there?  The original design concept was to save money and simplify operation, but I figured that since I built this transmitter specifically for use on forty meters, why not optimise it accordingly?  So.....I did.

There went my hopes for simplicity.  But I was glad I had room to make the change without seriously overcrowding things.  Here, you can see on the lower left a coil, wrapped around a 5/8" nylon form, that just happened to snug into a hole in the chassis.  On the lower right, you see a 100pF pilot variable, which again, fit into a convenient hole that already existed.  I only had to drill the two screw holes.  The coil as a computed value of 11uH.  The capacitor peaks the oscillator output at around 45-50pF.  

With the tight spaces getting tighter, I decided to take some parasitic precaution and added a 2.5mH RFC in the 6AG7 cathode line as well.  So now, my simple little Mini Bandmaster has three, count 'em, three RFCs.

That coax you see was problematic.  It added stray capacitance to the output.  It threw tuning off.  I removed it after this photo, replacing it with a single 20g. insulated wire, which the loading cap thinks is just dandy.  

When you build in tight spaces, especially with point to point wiring, be prepared for weird things like this.

So, I buttoned the little "Mini" up, and attached a fifteen watt light bulb to the output, and here is the result:


I thought it prudent to make a notation for my own records the changes I made to my "tried and true" little Novice MOPA.  This notation took the form of a revised schematic, which I share below.  A tip of the hat to the RSGB (Radio Society of Great Britain) regarding the schematic style.  These types of schematic stylings were commonly used in their publications of the 1950s and 60s, and I love them.  To me, they are ultra clear and easy to read.

I should add that I do not feel the RFC in the cathode of the oscillator is critical in value.  I would suggest anything from 500uH to 5mH.  If you really want to choke off parasitics, you could also place a 40 - 50 ohm one watt resistor in series just after the amplifier plate pin, winding about five turns of wire around it. This is actually a standard item in most higher power transmitters, but you rarely see it in these little rigs.  But it wouldn't hurt to include it if you have the room. 

So here it is, the completed NRR station project, as seen in the top photo.  I get a clean 15 watts output using the 6V6G, and 20 or more watts using the 6L6G.  The 6V6 dips at 70mA, the 6L6 dips at 90mA.  The signals are very clean.

So far, I have been pleasantly surprised to find that making contacts with this station is easier than I had thought.  The receiver is regenerative, and requires a spotting oscillator to locate the transmit frequency.  This is accomplished by that little box on the right side of the receiver.  It is a 2n2222a xtal colpitts oscillator.  I simply insert the transmitter xtal into it, turn it on, locate the heterodyne on the receiver, then remove the xtal and re-insert it into the transmitter xtal socket.  This places me well within proximity of anyone calling for me, on either side of zero beat.  The swamping of the receiver by the transmitter serves as a sort of "side-tone".  I also use one of the other receivers as a side tone, depending on my mood.  Nice to have choices, no?

This project has been completed for about a week since this writing.  I did use this set-up for NRR in February,  using the prior supply, generating five watts.  I still managed 360 points for NRR, and scored Sweden on 7045.  With all the improvements, I've scored both coasts of the US, a lot of Texas, but after midnight, I am getting calls from Europe!  Really!  Dx on forty meters has been really great lately.  F6HKA reported 559, and OK1FIM reported 579.   Huh.  Fifteen watt Mopa.  Regenerative Receiver.  Really.

This has been a most rewarding project.

I want to give special thanks to WB4WHH for his article, which was actually based upon my original article! His article can be found here.  I helped him with this project over the e-mails originally, several years ago.  I used his improvements on my original design as a guide.  Talk about a serendipity!  We've come full circle, OM!  :)

Well, here we are, the parting shot of the current operating position.  Left to right, Power Supply, Antenna Switch, Mini Bandmaster with SWR bridge atop, Regenerodyne, and spotting oscillator.  The key is a Royal Air Force 8 Amp from 1942, and the headphones are US Military low-Z from the early 1950s.  A Novice Station in 1930s clothing.

Now, all I need to do is print some new QSL cards!!

If anyone has further question, check out the Novice Rig Round Up Facebook Page.  I'll be there.  Also, I can be reached at

This station can be heard on the following frequencies usually around midnight EST on: 7035, 7037, 7045, and 7050.

'73, de wd4nka.

No comments:

Post a Comment