[minima] Re: experimental modified kiss mixer
- From: Ashhar Farhan <farhanbox@xxxxxxxxx>
- To: "minima@xxxxxxxxxxxxx" <minima@xxxxxxxxxxxxx>
- Date: Thu, 14 Aug 2014 06:40:23 +0530
Jon,
great work!
the topology is quite close to what we now call as the H-mode mixer.
though, this is the first time we are looking at an H-mode mixer with
discretes. As the four transistors are relatively unmatched (compared to a
switch like FST3157) we can't expect as good balance as is reported by
martein. i hope you have see that site. do you have a two signal IMD test
source? it would be fab if you could post some IIP3 figures so we can stack
this up against the H mode mixers using the FET switches.
for the minima, prevent leak of the local oscillator towards the IF is not
as critical as preventing the leak towards the RF port. The IF port has a
crystal filter that will automatically strip anything that is not within
it's passband. our issues have been only with the LO leakage towards the RF
port that might make it through the RF power chain during the transmit.
Those have largely been addressed now, I presume. There have been a few
discrepancies in the losses but they are probably due to some measurement
oversights. we must continue to confirm these.
my guess is that the leakage towards the IF port in all the KISS mixers is
due to the gate-drain capacitance of the FETs. they form a coupling from
the LO transformer to the mixing transformer. the J310 datasheet puts this
at 2pf. If this is indeed the cause, we must see a rise in the LO leakage
as we increase the LO frequency (the reactance of the gate-drain
capacitance will keep decreasing, leading to greater leakage). someone with
the specan could confirm this.
- f
On Wed, Aug 13, 2014 at 5:59 PM, Jon Lee G4TSN <g4tsn@xxxxxxxxx> wrote:
> Hi All
>
> Following on from yesterdays experiments
>
> And only getting the oscillator suppressed in one direction I have tried
> to modifiy the KISS mixer to get it to work in both directions, after much
> reading
>
> I have come up with a slightly less KISS approach which seems to be
> giving better results for me at least
>
> 2 extra fets and another transformer plus another pot
>
>
>
> With the standard si570 oscillator driving OSC port and -10dBm 20 MHz on
> either RF/IF port I see same results -18dBm out at the wanted frequency
> all other un wanted below 40MHz are -48dBm or below.
>
> Conversion loss looks ok at around 8dB ??
>
> The good thing is you can reverse the RF / IF ports and get almost
> identical results !
>
>
>
> The coils are all standard as per original minima mixer (note not quite
> drawn correctly here so make them like the original) all centre taps are
> left floating
>
> Please let me know any other ideas and comments, I am sure someone will
> improve on this
>
> 73,
>
> Jon G4TSN
>
>
Other related posts:
