EMRFD Message Archive 4644

Message Date From Subject
4644 2010-05-05 16:25:06 Stephen Wandling Building the Universal QRP Transmitter
While I am aware of the "Updated" Universal QRP Transmitter (UQT) I was
looking for something simpler and with lower parts count, as I am
building it for the Minimal Art Session (MAS) 'event' that's on in a
week. http://tinyurl.com/MAS-QRPedia

I have the EMRFD, but not the SDFRA, nor am I a QST subscriber, so I
don't have access to the original article. I have built a version I
found on the Internet from K0KP, without his keying circuit and with a
ceramic resonator VXO. The oscillator runs all the time I am in
Transmit and I'm keying the Vcc to the PA.

I don't see a lot of 'tank' circuits in schematics I run across these
days. Given the mA meter in the PA Vcc line in Sprat #52 and the 'rf
probe' circuit Wes put in the Updated version, I was presuming that the
intention was to peak the output with the tank capacitor in the
collector of the oscillator.

When I find the oscillator in the station receiver, I can peak its
output as measured at the dummy load on the output of the PA. But,
typically when I do this and then key the PA, there is no output. If I
then adjust the tank capacitor with the key down, I can usually get
output and peak it, but I hear the frequency shifting down on the
station receiver, often 4-6 Khz.

The other thing I am experiencing is that when the oscillator is
'peaked' on a frequency, I am unable to simply shift the VXO frequency,
as the range of oscillation is very narrow and it stops oscillating.

Question: Is there a public source of the original QST or SDFRA
article? I'm presuming that this would give me all the answers I need
regarding tuning this rig.

Thanks.

72
Stephen
VE7NSD
4645 2010-05-05 17:45:26 Stephen Wandling Re: Building the Universal QRP Transmitter
Sometimes, just having 'someone' to tell the problem to helps!! :-)

I should have also tried the rig with crystals instead of just with the
3.58Mhz ceramic resonator, before sending off my first post, below.
With crystals, both the HC-49 and the FT-243, it works quite well and is
stable. I'm getting one watt out with a beat up 2N3866 (a previous
project was great for letting hFE out of transistors).

So, while I would still be interested in the original article, any tips
on how to get it to function with the ceramic resonator would be
appreciated. The oscillator does function as a 60Khz range VXO with the
ceramic resonator, but I am unable to peak the tank to get a usable output.

72
Stephen
VE7NSD

Stephen Wandling wrote:
> While I am aware of the "Updated" Universal QRP Transmitter (UQT) I
> was looking for something simpler and with lower parts count, as I am
> building it for the Minimal Art Session (MAS) 'event' that's on in a
> week. http://tinyurl.com/MAS-QRPedia
>
> I have the EMRFD, but not the SDFRA, nor am I a QST subscriber, so I
> don't have access to the original article. I have built a version I
> found on the Internet from K0KP, without his keying circuit and with a
> ceramic resonator VXO. The oscillator runs all the time I am in
> Transmit and I'm keying the Vcc to the PA.
>
> I don't see a lot of 'tank' circuits in schematics I run across these
> days. Given the mA meter in the PA Vcc line in Sprat #52 and the 'rf
> probe' circuit Wes put in the Updated version, I was presuming that
> the intention was to peak the output with the tank capacitor in the
> collector of the oscillator.
>
> When I find the oscillator in the station receiver, I can peak its
> output as measured at the dummy load on the output of the PA. But,
> typically when I do this and then key the PA, there is no output. If
> I then adjust the tank capacitor with the key down, I can usually get
> output and peak it, but I hear the frequency shifting down on the
> station receiver, often 4-6 Khz.
>
> The other thing I am experiencing is that when the oscillator is
> 'peaked' on a frequency, I am unable to simply shift the VXO
> frequency, as the range of oscillation is very narrow and it stops
> oscillating.
>
> Question: Is there a public source of the original QST or SDFRA
> article? I'm presuming that this would give me all the answers I need
> regarding tuning this rig.
>
> Thanks.
>
> 72
> Stephen
> VE7NSD
>
4646 2010-05-05 18:30:20 Tim Re: Building the Universal QRP Transmitter
There is a very common phenomenon with oscillators, which I suspect is far more extreme with resonators. As you "peak" the collector tank in one direction you see a slow gradual rise and then a rapid fall. Turn it in the other direction and you see a rapid rise and a slow fall. But... the actual position of the peak output depends on which way you are tuning.

This is not a new phenomenon by any means... I think it has to do with the strong coupling between the output tank and the crystal or resonator. In fact the same thing happens (peaking in different places depending on tuning direction) with old-fashioned grid-dip meters and when it happens, it means that you are too tightly coupled. And when tuning the crystal oscillators in all the old Heathkits this phenomenon is mentioned too.

In any event, when you see this happening, the most stable place to peak the oscillator is on the slow side of the peak, not the fast side of the peak. The slow side is also the side that the oscillator will most reliably start at. Often it will not start on the fast side.

I see you debating a crystal vs a resonator... I will be very reasonable and suggest that for 80M you don't need either, a LC-tuned VFO can be quite stable and easy without going to heroic measures. If you don't have a fancy-pants dial drive, the bandset+bandspread scheme works perfectly well (and has the plus of adding more knobs!) It's a lot harder to build a stable VFO
4657 2010-05-06 11:43:18 Stephen Wandling Re: Building the Universal QRP Transmitter
Tim,

Thanks for the interesting information. I will play with my osc tank
tuning today and see it it matches your 'tuning pattern'.

I have been given a look at the original design from 1977 in Solid State
Design for the Radio Amateur. What I have built bears some resemblance,
but differs in some ways. One informative piece of info was that the
variable capacitor for the oscillator tank is specified as a trimmer.
The implication I get from this is that it can be adjusted and won't
need tweaking in the future? I guess this might only be valid if you
don't change crystal types, e.g., shifting from an FT-243 to an HC-49.

Your comments about more knobs as a bonus made me chuckle. I've built
this rig on a 'test bed' of 6" x 8" x 1/2" plywood, with a 3" high front
and back aluminum panels. My main circuit board is 2" x 4" PCB, screwed
to the wood base. This is home to my regen receiver and the
transmitter. The current number of knobs on the front panel is 7, but I
don't use the RF gain pot. I not only have bandspread for the receiver
tuning, but I have bandspread for the regen control! And actually, one
of the bandspread caps is 'outboard' of the front panel. There's no
more room for knobs!!

Thanks for the LC tuned VFO suggestion. If the MAS event wasn't one
week away, I would give it a try.

Thanks for your help and thoughts.

72
Stephen
VE7NSD

Tim wrote:
> There is a very common phenomenon with oscillators, which I suspect is far more extreme with resonators. As you "peak" the collector tank in one direction you see a slow gradual rise and then a rapid fall. Turn it in the other direction and you see a rapid rise and a slow fall. But... the actual position of the peak output depends on which way you are tuning.
>
> This is not a new phenomenon by any means... I think it has to do with the strong coupling between the output tank and the crystal or resonator. In fact the same thing happens (peaking in different places depending on tuning direction) with old-fashioned grid-dip meters and when it happens, it means that you are too tightly coupled. And when tuning the crystal oscillators in all the old Heathkits this phenomenon is mentioned too.
>
> In any event, when you see this happening, the most stable place to peak the oscillator is on the slow side of the peak, not the fast side of the peak. The slow side is also the side that the oscillator will most reliably start at. Often it will not start on the fast side.
>
> I see you debating a crystal vs a resonator... I will be very reasonable and suggest that for 80M you don't need either, a LC-tuned VFO can be quite stable and easy without going to heroic measures. If you don't have a fancy-pants dial drive, the bandset+bandspread scheme works perfectly well (and has the plus of adding more knobs!) It's a lot harder to build a stable VFO