[sdiy] passive ring modulator transformers

[Magnus Danielson]

Hi,

On 01/17/2016 04:48 PM, mskala at ansuz.sooke.bc.ca wrote:
> On Sun, 17 Jan 2016, Magnus Danielson wrote:
>> In another context, we concluded that there seems to be around 30 dB better
>> S/N possible with diode mixers compared to the Gilbert-cell mixer of MC1496
>
>> While the experience was in the RF field, basic properties should translate
>> down to audio frequencies.
>
> Is that with or without harmonics?  RF applications tend to be rather
> narrow-band, so that harmonics from the diode switching, of which there
> are a lot, will end up getting filtered and probably not counted towards
> the S/N ratio.  But the original poster here was hoping for three decades
> of bandwidth, and then it matters whether you count the harmonics from the
> switching as part of the signal or part of the noise, either of which
> might make sense depending on what you actually want in the output.
>

Good point.

However, the same basic problem occurs in both worlds. As you drive the 
Gilbert cell stronger, to too will experience more harmonics. In the end 
of the day, it's the same PN junction in actions for both cases.

Cheers,
Magnus