My First-Ever QSO -- 50 Years After Passing Ham Test

Discussion in 'Ham Radio Discussions' started by G3EDM, Aug 27, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-Geochron
ad: Left-3
ad: l-BCInc
ad: Left-2
ad: L-MFJ
ad: abrind-2
  1. G3EDM

    G3EDM Ham Member QRZ Page

    Concerning that CW filter, I now have the info from someone who knows about these things:

    "It’s the filter to meet the 'Very Narrow Requirement for HF Maritime Receivers' to meet the UK Spec MPT1204 for HF receivers compulsorily fitted in ships. That's for narrow CW. The basic BP4725 was 600 ohms and the -01 is not in my data, but is almost certainly either 50 ohms or 1kOhm in shunt with 15pF."

    73 de Martin, G3EDM
    G0CIQ likes this.
  2. SM0AOM

    SM0AOM Ham Member QRZ Page

    To use such a narrow filter in a homebrew receiver requires a good dial without backlash,
    a slow tuning rate, less than 25 kHz/turn is advisable. Also, the frequency stability needs to be good, so your receiver returns to the last dialled frequency after each transmitting session.

    I had some correspondence in the 70s with SM5MN(SK), an experienced receiver builder who made a few custom A1 receivers for VHF weak-signal work. He claimed that the narrowest selectivity that was practically usable was in the order of 200 Hz, and only with an extremely stable receiver and a very good dial.

    It takes some time to be accustomed to an A1 receiver with a multi-pole IF filter with
    good skirt selectivity, and some operators do not like the tuning behaviour of such a receiver at all.

    Next question is how a receiver should be built to make full use of such fine filters.

    There is a lot to be said in favour of distributed selectivity, with a wider filter in front of an IF strip where another, narrower, filter is switched in downstream when required. This of course requires that all filters have the same centre frequencies.

    Years ago, the KVG XF-9 series offered several bandwidth choices having 9 MHz centre frequency,

    Some contradictory requirements exist in such designs, and a lot of experience is needed to make good engineering design considerations. Suggest that you read the works of W1FB(SK), W7ZOI and W0IYH(SK) for some pointers to good design practices.

    Also G3VA(SK) collected a lot of receiver design wisdom in his "Technical Topics" columns, which are available in condensed form.

    Last edited: Sep 20, 2021
    G0CIQ likes this.
  3. G3EDM

    G3EDM Ham Member QRZ Page

    More housekeeping. Here's the schematic, and voltages, after adding a "voltage regulator" tube to the transmitter oscillator's B+ supply today. Full writeup coming soon (on the Homebrew forum) and a summary will be posted here too.


    By the way if you are drawing your circuits by hand and not using drawing software, these plastic templates are great:


    73 de Martin, G3EDM
    KA4DPO likes this.
  4. G3EDM

    G3EDM Ham Member QRZ Page

    Some progress again.

    This is an abbreviated version of a longer post in the Homebrew forum. For the full technical details, go there.... the thread is called "Oscillator/PA Cathode Keying, 'Spotting' Your Frequency, and Chirp" and it's post #109.

    Today I added an OA2 voltage regulator tube to the oscillator plate supply. Once again this was kitchen-counter electronics. I wired the OA2 "outboard" to make the modification easily reversible if necessary. The lovely blue/purple glow of a VR tube:


    At the plate of the 6AB4 oscillator, I measured 142V (prior to adding the VR tube it was 233V!).

    Tests of the transmitter after adding the VR had these results:
    • To get consistent oscillation on all the crystals, I had to raise the value of the variable resistor that is wired across R1.
    • All the crystals oscillated, even the ones that did not oscillate before (that puzzles me greatly, given that drive is now lower, but never mind).
    • All the crystals now sound very much the same; some are better than others but the difference is no longer dramatic.
    • There is considerably less "yoop" (change in the tone immediately after key-down).
    • The remaining defect in the signal is a substantial ripple in the tone, above and below the centre frequency. I personally find it quite distracting but I think such a signal can be copied fine by other hams (albeit with some irritation!) and will not require constant twiddling of the RIT.
    • I am very suspicious of my regenerative receiver's quality in monitoring this ripple. Simply touching the receiver, or gently tapping the table, results in an identical audible "wobble". So I now think it would be worth doing a controlled test with the help of another local ham and a "known-good" receiver.
    • I am assuming that with the much lower plate voltage, my crystals are now far less prone to heating, cracking, and damage in general.
    This modification is a "keeper". I now have to find space inside the chassis for that tube and build a mounting bracket.

    Here is a recording of my keyed signal after adding the VR tube:

    (For the latest schematic, see post #823.)

    73 de Martin, G3EDM
    Last edited: Sep 21, 2021
  5. W9BRD

    W9BRD Ham Member QRZ Page

    For general two-way comms with a regen, the main need is low-pass AF filtering, as regens are generally wide. I use 10-k input/output impedance LMI-series LPFs by UTC/Triad. Search at Ebay on "UTC filter", maybe "UTC filter LMI", and you'll see what's currently there. Anything over $20 US for one is scalping; just laugh such sellers off. Even Surplus Sales of Nebraska ( ) sells brand-new LMI-1000s for $17.95 US. How I use them is shown in the circuit at the bottom of . Turns out that the sound of even military and commercial receivers (R-4A, BC-348) can be improved with the AF amplfiier/filter box and its low-pass filtering.

    And BTW, I also prefer low receiving pitches -- generally 500 Hz or less. Try that with a built-from-a-magazine-article "heyday of the regen" regen that pulls and overloads like the dickens...
  6. W9BRD

    W9BRD Ham Member QRZ Page

    You're already exploring the effect of underdrive as a result of not only stabilizing, but also reducing, the oscillator plate voltage by means of the 0A2. (BTW, in keeping with being "That Guy": the first character in that tube type/designator is a zero, not an oh. It's there because a 0A2, best pronounced zero A 2, has a heating voltage of 0, just as a 6AQ5's type/designator starts with a 6 because we heat the tube with 6.3 V.) Although the crystal strain will be lessened and signal quality will be improved as the oscillator plate voltage is reduced, my main aim in suggesting the VR-tube addition was stabilization of the oscillator plate voltage with keying and not reduction of oscillator plate voltage.

    To keep the same drive level, you'll need to keep the same oscillator plate voltage; for instance, seriesing an 0A2 (150 V) and a 0C2 (75 V) will give a regulated voltage quite close to what you measured before adding the VR tube. But whether or not that's doable depends on whether or not the unregulated, full output of your power supply is sufficiently high to allow a 225-V VR string to strike (initially light) and then stay lit when the power-supply voltage drops in the KEY DOWN state.

    This is all excellent learning experience. But it's also why the Boosted Pierce is a "modification pit" (after the popular "money pit") -- sufficiently better performance always seems like yet one more incremental improvement away...

    ...except anything like modern signal quality with unstressed HC-49 rocks isn't because of the built-in limitations of the BP and the physical tininess of HC-49s. But "good enough for your purposes" may just be achievable, especially at 80 meters.

    The answer to getting modern-signal quality with HC-49s and vacuum tubes at the tens-of-watts level is a three-stage transmitter. I'll get my general circuit for same, developed over multiple years of experimentation and proof, drawn up "pretty soon." To keep HC-49s safe, we must operate any oscillator based on one at a sufficiently low level -- a lower level than any classical Novice-power-level-class transmitter circuit (with the conditional exception of the three-stage Drake 2-NT) will afford. It boils down merely to needing a transmitter lineup with higher gain from oscillator grid to ANTENNA jack than any classical two-stage circuit and its RF-power-providing oscillator can provide.
    Last edited: Sep 21, 2021
    N2EY and G3EDM like this.
  7. HA2ZB

    HA2ZB XML Subscriber QRZ Page

    This topic evolved to a top-class technical discussion about crystals, valves and related design and construction aspects. I love reading this, although I am afraid I do not have the skills to build similar equipment - so this is out of scope for me unfortunately.

    However, I guess it would make sense storing (linking?) this thread under the Technical Forum as well (General? Homebrew?) so later it can be easily found by someone who was not here from the beginning. Keywords could be added too, and the copy may have a more descriptive title. (No problem with the current title at all, but I personally often search in titles only.)

    Just an idea, to keep this content available for people who have similar projects in mind and looking for guidance.
    G0CIQ and W9BRD like this.
  8. KP4SX

    KP4SX Premium Subscriber QRZ Page
  9. HA2ZB

    HA2ZB XML Subscriber QRZ Page

  10. G3EDM

    G3EDM Ham Member QRZ Page

    Just to make sure I completely understand, you are saying that although reducing and stabilizing the oscillator voltage seems to have been gentler on my crystals, the combo will not work because the stabilized oscillator plate voltage (of around 150V) is too low for sufficient drive to the 5763?

    Is there a combo of voltage stabilizer tubes that would get us, say, more like 200V? I would have thought there must be. In that case I would hope my PSU is capable to providing enough overhead above that voltage to allow the VR(s) to operate correctly (prior to installing the VR, the voltage delivered by the voltage divider was 233V). Also I assume we could play with the value of the series resistor on the VR(s).

    But we still do not know for sure whether I have a problem with sufficient PA drive. If you look at the thread in Homebrew you will see that after installing the VR, I initially completely mis-tuned the TX. Once I tried a "lightbulb dummy load" instead I was able to get something like normal drive by tuning to maximum brightness.

    So among the many tasks on the to-do list is to wire in that 100-ohm resistor and see what voltage drop we get, finding out what drive we actually have.

    Also, frankly I still don't really know what my signal sounds like on-air. I just don't trust that regenerative receiver for that kind of thing. That needs to be resolved because otherwise, we will continue to get a series of apparent improvements, followed by disappointments, caused in part by lack of reliable data on the signal.

    A lot of progress has been made, for sure....

    73 de Martin, G3EDM

Share This Page