Hi Markus,

Mine too! I would strongly prefer that every exciter and amplifier
be linear but sadly that is not the case in my station. The other
monster in the chase is frequency stability. ;-)

I do have a SSB HF transceiver and a home built downconverter for LF
transmitting. I used this for some CW and JT9-1 QSOs. Unfortunately
it is of no use for slow modes. The converter has a 10 MHz OCXO
which is very stable but the HF transceiver operating at 10.136 MHz
has a cyclic thermal drift of about 6 or 7 Hz every time my furnace
comes on and changes the room temperature! Transmitting, even at low
power will also cause problems with heat. This is no good at all. I
could try to put some kind of heater (like a crystal oven) on the
transceiver's TCXO and try to add some insulation around it. This
would not be easy due to its location on a large PCB and limited
space. I doubt I could make it good enough.

Better of course (and very interesting!) would be to replace the
TCXO with an external GPS locked oscillator but I do not know if I
am up to this challenge. It would require some very delicate surgery
I guess.

In an ideal world the HF transceiver TCXO would be replaced by a
GPSDO, the 10 MHz OCXO in the downcoverter replaced by a GPSDO, the
downconverter rebuilt to be linear, the amplifier replaced with one
that is linear, and the LO in the SDR receiver replaced by a GPSDO.
Unfortunately we do not live in an ideal world and probably none of
that is going to happen. :-)

73,
Paul



On 11/16/18 9:18 AM, Markus Vester wrote:
> Hi Paul,
>
> guess my preferred method is a linear SSB exciter which can do
> all sorts of modes, including specialties like chirped Hell and
> picture transmissions, or 1-dB stepped power tests. On the other
> hand with one sitting here this is easily said ;-)
>
> Stefan and Tom have successfully employed a Raspberry Pi, which
> can produce modulated RF directly on one of it's digital output
> pins. This has been used for WSPR but I guess it could also make
> DFCW.
>
> Best 73, Markus (DF6NM)
>
>
> -----Ursprüngliche Mitteilung----- Von: N1BUG <[email protected]>
> An: rsgb_lf_group <[email protected]> Cc: lowfer
> <[email protected]> Verschickt: Fr, 16. Nov. 2018 12:34
> Betreff: Re: LF: Methods to transmit DFCW
>
> Thanks Andy. I have no proficiency at all in PIC or Arduino
> programming. In my younger days I wrote some reasonably complex
> DOS programs but a lot has happened since then an the old brain
> isn't what it used to be I'm afraid.
>
> I am not against *trying* to learn but will consider this a last
> resort if no more complete or operator friendly system can be
> found.
>
> 73, Paul
>
>
> On 11/14/18 10:42 AM, Andy Talbot wrote:
>> Depends how proficient you are with PIC or Arduino or Whatever
>> programming. There are still AD9850 moduiles to be found on
>> Ebay - that are not completely obsolete or sold-out yet.
>>
>> Write your controller code to switch between two or three (for
>> DFCWi) frequencies under control of the ON7YD software,
>> toggling RTS / DTR on a COM port interface (use an FTDI USB one
>> if necessary)
>>
>>
>> www.g4jnt.com
>>
>>
>>
>> On Wed, 14 Nov 2018 at 11:50, N1BUG <[email protected]> wrote:
>>
>>> I am looking for a better method to transmit DFCW for 2200
>>> meter QSOs. For my QSO with 2E0ILY I used a QRP Labs U3S for
>>> the transmitter. This worked very well for DFCW60 when I was
>>> transmitting every other night. For any QSO attempts
>>> involving quick change from RX to TX it would be difficult
>>> because of the time required to program a new message!
>>>
>>> One possibility would be to modify this method used by N1VF
>>> for JT9 QSOs with a U3S:
>>>
>>> https://bgelb.github.io/u3s-qso.html
>>>
>>> I am a little concerned about messing up my trusty U3S but I
>>> may try that.
>>>
>>> It looks like the easy way is the old QRS software by ON7YD.
>>> This is very interesting but I don't have a suitable
>>> transmitter. Are there any published designs for DFCW
>>> transmitters which could be controlled in this way? I wonder
>>> if parts would still be available? I am guessing these used
>>> synthesizers which are not current generation any more. :-)
>>>
>>> Any other suggestions?
>>>
>>> 73, Paul N1BUG