[time-nuts] About HP10544A
Bruce Griffiths
bruce.griffiths at xtra.co.nz
Sun Mar 8 01:12:40 UTC 2009
Esa Heikkinen wrote:
> > Its probably better to increase the emitter current to 40mA or
> > so and place 47 ohms between the emitter and the output coupling
> > capacitor a and accept the resultant +7dBm output.
> > The transistor dissipation will increase and it may be better to
> > substitute something like a 2N5943, 2N3866, 2N 5109, BFW16A
> > or similar.
>
> I don't have those 2N types but after some digging found 2N2219A.
> Changed that and 47 ohms from emitter to the ground. I run it first with
> 6 volts, then with 9 volts and then with 12 volts but have to stop with
> uncomplete sweep because it become very hot.
>
>
Thats why I suggested transistors in T05 or T039 cases so that you can
use a heatsink.
> Increasing the current does help with amplifier linearity but not so
> much. However the spectrum looks little bit better now but it's still
> out of spec:
>
> http://www.amigazone.fi/files/gpsdo/544-7.png
> http://www.amigazone.fi/files/gpsdo/544-8.png
> http://www.amigazone.fi/files/gpsdo/544-9.png
>
>
>> About the only way to achieve the distortion you observe is if the
>> capacitance of the wire connecting your emitter follower to the 10544A
>> output is too large (around 50pF or more).
>>
>
> It's about 30 cm. long microwave coax taken out from some RF stuff
> (maybe old cellphone). It's measured capacitance (at the other end free)
> is 36,2 pF.
>
> I'll send the picture later if still needed..
>
>
Esa
36pF from the cable plus a little additional stray capacitance is
probably more than sufficient to cause most of the observed distortion.
Try using very short leads (< 5cm long) to connect to the emitter
follower buffer.
You need to keep the capacitance < 10pF or so.
Bruce
More information about the Time-nuts_lists.febo.com
mailing list