Message	Date	From	Subject
7109	2011-12-23 14:50:25	Chris Trask	Mixer Musings and the KISS Mixer (revised)
I just completed revising the paper on the KISS mixer and have added it to my web page rather than update the group Files version, which I will be removing shortly. The revised version can be obtained at: http://www.home.earthlink.net/~christrask/Paper026.html Chris Trask N7ZWY / WDX3HLB Senior Member IEEE http://www.home.earthlink.net/~christrask/
7110	2011-12-24 19:35:48	kb1gmx	Re: Mixer Musings and the KISS Mixer (revised)

7111	2011-12-25 07:00:36	Chris Trask	Re: Mixer Musings and the KISS Mixer (revised)
> > > I just completed revising the paper on the KISS mixer > > and have added it to my web page rather than update the > > group Files version, which I will be removing shortly. > > The revised version can be obtained at: > > > Very good. Its' certainly simple and parts frugal as well. > > I did a bit of study and noted that the evolution of the > KISS mixer is long. The first variant of a single balanced > switching mixer was Rohde using 2N5179 transistors for the > switch. Later I would see designs by Maas and Campbell > using FETs as the switch device. Now with a cheap and > inexpensive IC you have simplified their designs by > eliminating or more correctly using the ICs integrated > device drive where previous designs were done discreetly. > And I've had a few more passed along to me since I first posted this. Seems that the idea had been bouncing around in the ether for some time. > > What I wonder for KISS is: > > Load sensitivity? > Not tested yet, though I expect it will be fairly good since there is very little in the way of nonlinearity sources in the circuit. The biggest one may be the transformer core, and that will need to be determined by testing. > > Port to port isolation? For single balanced I'd > expect RF to IF to be good but LO to any is what? > It's not obvious, but it is double balanced since the there is no LO voltage actoss the transformer secondary windings. The LO/RF and LO/IF leakages are not good at the moment, and I need to see if adding a resistor from the Vcc/2 source to the transformer secondary centre tap will make a difference. > > Any issues around transformer design such as bandwidth or core > limitations contributing to IMD? > That's going to be a big issue here. The transformer I used is a design that I use for small-signal HF circuitry. For VHF I will switch to -61 material and maybe a shorter core, specifically the Fair-Rite 2861002702, or perhaps even powdered iron cores. For low frequencies I'll use the Fair-Rite 2873002402. But the core size is also up for scrutiny. My prototype ia already falling apart. If anyone is interested I could design a small single-sided PCB with the mixer and driver circuitry with pads for an MCL transformer and add it to the paper. Chris Regards, Chris

>
> > I just completed revising the paper on the KISS mixer
> > and have added it to my web page rather than update the
> > group Files version, which I will be removing shortly.
> > The revised version can be obtained at:
>
>
> Very good. Its' certainly simple and parts frugal as well.
>
> I did a bit of study and noted that the evolution of the
> KISS mixer is long. The first variant of a single balanced
> switching mixer was Rohde using 2N5179 transistors for the
> switch. Later I would see designs by Maas and Campbell
> using FETs as the switch device. Now with a cheap and
> inexpensive IC you have simplified their designs by
> eliminating or more correctly using the ICs integrated
> device drive where previous designs were done discreetly.
>

And I've had a few more passed along to me since I first posted this. Seems that the idea had been bouncing around in the ether for some time.

>
> What I wonder for KISS is:
>
> Load sensitivity?
>

Not tested yet, though I expect it will be fairly good since there is very little in the way of nonlinearity sources in the circuit. The biggest one may be the transformer core, and that will need to be determined by testing.

>
> Port to port isolation? For single balanced I'd
> expect RF to IF to be good but LO to any is what?
>

It's not obvious, but it is double balanced since the there is no LO voltage actoss the transformer secondary windings. The LO/RF and LO/IF leakages are not good at the moment, and I need to see if adding a resistor from the Vcc/2 source to the transformer secondary centre tap will make a difference.

>
> Any issues around transformer design such as bandwidth or core
> limitations contributing to IMD?
>

That's going to be a big issue here. The transformer I used is a design that I use for small-signal HF circuitry. For VHF I will switch to -61 material and maybe a shorter core, specifically the Fair-Rite 2861002702, or perhaps even powdered iron cores. For low frequencies I'll use the Fair-Rite 2873002402. But the core size is also up for scrutiny.

My prototype ia already falling apart. If anyone is interested I could design a small single-sided PCB with the mixer and driver circuitry with pads for an MCL transformer and add it to the paper.

Chris

Regards,
Chris