[ARC5] ARC-5 transmitter frequency drift suggestion

[Dennis Monticelli]

Freq drift in the 40M units is the most problematic of course.  I run
a T22 with a voltage regulation of the osc which in turn runs
continuously while the PA cathodes are keyed.  After a warm-up period
it is quite stable over the course of a long QSO.  Chirp is another
matter.  After doing the obvious (clean all contacts in the signal
path, new HV bypass caps, etc), the next step is to regulate the
screens of the PA.  This definitely helps and its easy to do.  But to
eliminate the chirp altogether requires that the HV to the PA plates
be regulated also.  I have switches on my power supply that allow me
to independently activate/deactivate regulation of the screens and
plates of the PA.  The cause and effect is quite apparent.  I don't
run a cathode resistor in the PA because it robs some output power.
With 500V HV, Pin is 50W and Pout is 31W.  That was enough to for 40M
DXCC using my dipole.

BTW, do take care of the clicks (standard fix) while addressing the
drift issue.  Your fellow hams will appreciate that.

Dennis AE6C

On Sat, Dec 6, 2008 at 11:24 PM, Dennis Monticelli
<dennis.monticelli at gmail.com> wrote:
> In order for a PA to be driven to 100% AM, the drive to the input must
> be sufficient to support 4X peak power.  In other words, it must be
> overdriven.  I suppose if the PA is grossly over-sized for the power
> desired, then AB1 would represent sufficient drive.  Otherwise, Class
> C running full rated grid current is warranted.
>
> Dennis AE6C
>
> On Sat, Dec 6, 2008 at 11:17 PM, Neil <neilb at ihug.co.nz> wrote:
>> Frequency drift in the ARC-5 and SCR-274N transmitters can be mostly
>> attributed to the 1626 oscillator being run at high power level in order to
>> supply adequate drive to the 1625 grids.
>>
>> It occurred to me some time ago that if the 1625's were to be run in Class
>> AB1 instead of Class C, the drive power required would fall to virtually
>> zero. This could be easily done by lifting the grounded end of the 1625
>> grid resistor (15K) and feeding -35V bias to it.
>>
>> The 1626 would now be required to supply a maximum of 35 volts of drive
>> (at virtually no power), versus the previous 100 volts or so at about 7mA
>> (half a watt). Its plate voltage could be scaled back accordingly, to
>> perhaps
>> 150 or even 105 VDC regulated.
>>
>> I'm sure this would work very well in the case of CW, but I can't recall
>> anything in the literature describing AM plate modulation of a Class AB1
>> power amplifier. Has anyone tried this or can point to literature on the
>> subject?
>>
>> 73 de Neil ZL1ANM
>>
>> _______________________________________________
>> ARC5 mailing list
>> ARC5 at mailman.qth.net
>> http://mailman.qth.net/mailman/listinfo/arc5
>>
>>
>