Hi,
So the square vave drive to a bus switch must saturate better than the
SD5000. Racal drives the gates of the SD5000 through a 1:2 transformer with
about 10 volts peak to peak drive.
This means the gate pairs go about 10 volts peak to peak. I was wondering if
you took a AC bus driver one that is 8 bits and split them in half in two pairs
of 4 drivers. Then AC couple it to a transformer in bridge mode to get 10 volts
peak to peak drive. I'm considering converting the first mixer of my RA6830
from a ring to H but presently the weak link appears the roofing filter not the
mixer so may have to do a homebrew module with a pair of roofing filters. The
Racal mixer can take a HP8640b at full bore and appears to be just starting to
saturate at full bore of plus 19 dBM.input.
The bus switch sure works well in my SDR module I can get 96 dB of dynamic
range but it was limited by the sound card not my input board. frank
Giancarlo <[EMAIL PROTECTED]> wrote:
Hi Frank,
thanks for clarifying your question
--- In [email protected], FRANCIS CARCIA <[EMAIL PROTECTED]> wrote:
>
> Hi,
> I was thinking about discrete switching devices like a good fast
transistor that operates at a lower base voltage rather than logic
levels.
I do not know but you need good fast switching transistors in a
single chip. Probably you will need power to drive them and maybe
they will not have a low Ron... maybe some other problems too.
> Another thing I was considering was going back to using an SD5000
with two AC cmos gates driving a transformer in bridge mode to get
higher gate swing rather than 7 volts on a flip flop. frank
Well you can try .. the SD5000 was replaced by the FST3125 because it
was needing a very high switch on gating voltage, a substrate voltage
and it had a very had conversion loss over 8dB plus complex gating
balance adjustment. Not a hot front end ...
FST3125 gives the H-Mode Mixer (and others) a KISS solution, simple
driving no adjustment except squarer balance unless IP3 around
>+35dBm vs 40+ is OK you can avoid balance adjustment, and vy
important a very low conversion losses <5dB !!!! and lower noise than
SD5000 solution.
Finding a better solution is welcomed ... that's why we should share
ideas ....
The QSD did get a benefit by my idea of using Fast Bus Switches for
mixers.
73
Gian
I7SWX
> Giancarlo <[EMAIL PROTECTED]> wrote:
> --- In [email protected], FRANCIS CARCIA
<carcia@> wrote:
> >
> > Anybody ever try using 2N5109s in an Hmode mixer with RF drive.
> Frank WA1GFZ
> >
>
> Hi Francis,
>
> I am not sure I understood your question... do you mean if anybody
> has tried an RF amp at the H-Mode Mixer RF input?
>
> If this is the question.... an H-Mode Mixer is a Hot device (vy low
> conversion losses <=5dB. In low HF bands you do not need preamp as
> the requirements vs band noise are strong to IMD. In high HF bands
> you may be OK with a preamp where you may require also sensitivity
> and you can be OK dropping the IP3 from +40dBm(+/-) to >+30dBm.
> A low noise lossless feedback RF amp using 2N5109 should be OK.
Keep
> gain not higher than 10dB and with the highest IP3 possible.
> I have an low noise RF preamp in my overmodified IC751 with a 14dB
> gain replacing the original Icom pre- using 2x BFR91 in parallel.
No
> problems... I usually keep it off.
>
> 73
>
> Gian
> I7SWX
>
