A fix for drift?

Does anyone else have a slow drift problem with their Hull Hartley? I was looking at mine and spotted something I wanted to check into.

The center tap resistors for the filament are mounted right under the tuning cap. Any heat they generate goes right to the tuning cap. Hull suggests 50-200 ohm resistors there. I used a couple of 40 ohm resistors thinking that as long as my power supply can handle the extra current draw I'd be ok. Now I have a .7 watt heat source where Hull's design had as low as a .14 watt heat source there. Any heat might cause a problem. oops.

These two resistors are in the circuit to provide a ground point for cathode keying, but there is an alternative. I have a center tapped 7.5 V filament supply. I can remove those two 40 ohm resistors and just key the filament transformer center tap. The heat source will be removed so my Hartley should now stay on frequency better.

*** clock ticks while solder smoke rises ****

Well, I removed the 40 ohm filament center tap resistors from my Hull Hartley. It is a fairly clean change. The only visible indication of the modification is the addition of one fahnestock clip along the left side for the transformer center tap connection. Now that heat source is gone.

I did before and after standby drift measurements. After tuning on the rig I checked the transmit frequency every 30 seconds for 15 minutes. Key down was only long enough every 30 seconds to get a transmit frequency readout. I had no load connected and the plate current was about 18mA with 300 volts B+.

Was is worth it?

*** Now drums rollllll ***

It depends

On 80 meters I found that the original Hull design drifted about 2KHz from a cold start while the modified version drifted about 500Hz. That's good but both took about 15 minutes to settle down. With both versions I still have to be careful about jumping on the air right after power on. I saw no significant change on 40.

But two parts were eliminated along with some amount of power on drift, both good things.

If I were starting out now building my Hull Hartley and I had a center tapped filament transformer I'd make this change. If I already had my Hull Hartley up and running, I'd probably let it be.

Push-Pull Colpitts Transmitter - 8

I've had my Push-Pull Colpitts on the air now for about 3 weeks and 10 QSOs. Notable QSOs have included my friend Lou, VE3AWA, in Ontario and W0NYQ, a fellow Minnesotan with a big interest in '29 rigs.

This tranmitter is a keeper. Why?

- Efficiency is 50-60% (great for a '29 power oscillator)
- Drift is measured in the 10s of hertz range. Essentially solid as a rock.
- No hand capacity problem when tuning to get on frequency
- Easy to adjust feedback for minimum plate current and best performance
- "Perfect sounding 1929 signal"

I did receive a report of a trashy buzz on my sidebands. Since then I've been more careful when adjusting the feedback and also the output coupling. I haven't had any other bad reports.

I do still have some trouble with FMing/wobble due to the wind blowing around my antenna.  This is a common problem with '29 power oscillators. Whether this is an improvement over my other '29 transmitters, I can't tell.

I haven't cleaned it up on 40 but I can work on that later.

All in all, a really satisfying project.

CW on 3565 KHz

Push-pull Colpitts power oscillator on 3565KHz with 6 watts into a low endfed wire as copied by VE7SL in British Columbia, 1490 miles to the west:

It's on the air!

It is on the air! This evening I had a great 80 mtr QSO with my friend Lou, VE3AWA, in Ontario, 570 miles as the crow flies. He reported a perfect sounding '29 signal. My Push-Pull Colpitts Power Oscillator is also amazingly stable. Even from a cold start it is on frequency.

I plan to leave it set up next to my Novice Rig Roundup operating position so that I can put it on the air whenever I want to.

Music to my ears

It works!

Last night I completed the under-the-chassis wiring of my Push Pull Colpitts transmitter. I ended up doing "bus bar" wiring. It gives a 3D effect and worked well for this design/layout. Between that and the wooden chassis this one feels like a '29 transmitter. Most of the under the chassis parts are 40s vintage but, hey, they will be hidden.

 This morning I connected power for an initial checkout. With 300V on the plates and about 20 watts input I'm getting over 10 watts out of sweet sounding RF on 80. The grounded tuning/tank capacitor effectively eliminates the detuning effect of hand capacity and I'm certainly not going to complain about 50% efficiency. Open circuit voltage across the key terminals is only 48 volts so that's good too.

 Next I'll clean up the operating position and finish my 40 mtr coils so that I can test it there.

Push-Pull Colpitts Transmitter - 4

In between shoveling snow and chipping ice here in Minnesota (a common winter pastime and good way to catch up with the neighbors) I've continued to make progress on my push-pull Colpitts oscillator/transmitter. I've now got all of the above board parts mounted and wired  (except for the filament and B+ connections).  It is looking pretty impressive.

Along the way I've found that Lowes here in Rochester has the best selection of brass nuts, bolts and screws while Albany County Fasteners  has the best selection of specialty brass hardware like knurled nuts and wing nuts.

Next I need to plan the layout of the small parts that mount underneath.

Push-Pull Colpitts Transmitter - 3

I'm making good progress on my new transmitter.  With a lot of help from my friend KE0EXE and his table saw I now have a nice looking oak base. I went ahead and shellacked it so that it looks pretty good. 3/16" copper tubing for the antenna link and 80 meter coil was not available locally so I have to go to Amazon for that but I did find 1/4" tubing for the 40 mtr coil at my local big box hardware store. I added a National type A vernier dial to the Cardwell tuning capacitor. This should help a lot getting on frequency.

The next task is to find all of the screws, nuts and bolts that I need and then start fastening everything in place.

Push-Pull Colpitts Transmitter - 2

I think I have enough to get started....

The tough part to find was the dual 500 pf Cardwell capacitor. I found one by putting my need out on the AWA Group Yahoo group. What I now have isn't exactly what QST called for but should be close enough. A trip to Menards got me wood for a base. I'll need to pay a visit to a friend with a table saw. Copper tuning I should be able to find locally at Menards, Lowes or Home Depot. The small

parts under the "chassis" I should have on the shelf.

Next step is to cut and varnish the base.

Push-Pull Colpitts Transmitter - 1

I’ve started on a new transmitter for the next AWA Bruce Kelley 1929 QSO Party. The “BK” requires non-crystal-controlled transmitters that are of 1920s design and use tube types that were available in 1929. I found a transmitter described in the January 1934 issue of QST that meets these requirements.

For a good over view of the transmitters built to be used in the BK take a look at VE7SL’s gallery page at https://qsl.net/ve7sl/29gallery.html . Of the 65 transmitters 48 are either Hartleys or TNTs. The next one in the  list was the TPTG with 9. Only 2 were Colpitts. I have TNT and Hartley transmitters for the BK and a TPTG is a lot like a TNT so building a Colpitts for the BK sounds like a good next choice. This will round out my 1920s “big three”.

The 1934 QST article lists several features of this design that sound attractive:

Antenna coupling to the non-plate portion of the tank coil. The more common link coupling to the plate ends of the tank coil results in more second harmonic output and less frequency stability when the antenna is tightly coupled for maximum output. Center/swinging link coupling should be an improvement.

Push-pull tube capacitance in series. Temperature changes within the tubes that impact the interelectrode capacities will have less impact on frequency stability. This configuration also reduces the amount of current through the tubes.

Symmetric layout. A symmetric layout is less prone to exhibit signal instability

Grounded tuning capacitor body. In this design the tank tuning capacitor body provides shielding so that hand capacity has less of an impact on frequency. This transmitter should be easier to get on frequency.

Easily converted to an amplifier. If I decide to move on to a MOPA BK rig this will be one section that I already have.

In addition to information about this particular transmitter design, this QST article also gives hints that are useful for any 20s transmitter:

Use a well-regulated/stiff power supply

Route power and antenna cables away from the transmitter

Do not place the power supply close to the transmitter

Now, may the parts hunt begin!

Handbooks on the web

Several ARRL (and Radio) Handbooks can be downloaded from the web ...

1976:  https://ia800401.us.archive.org/27/items/RadioAmateurHandbook1976/radio-amateur-handbook-1976.pdf

1941:  https://www.scribd.com/document/91745963/The-Radio-Amateur-s-Handbook-1941

1936:  http://www.smcelectronics.com/DOWNLOADS/1936%20ARRL%20HANDBOOK.pdf

1931:  https://ia802807.us.archive.org/6/items/Amateur_Radio_Handbok_-_ARRL_1931/Amateur_Radio_Handbok_-_ARRL_1931.pdf

Many years of the "Radio Handbook" are at https://www.americanradiohistory.com/Radio_Handbook.htm

For ham radio magazines including QST, CQ and Ham Radio go to americanradiohistory.com .  In the upper right corner click on DX logs – SW Ham.  From there you will see many many classic radio magazines that you can download.