From: Roger Traylor (email@example.com)
Paul and Gang,
You stated that....
> ....... As a diode turns on, it makes noise as the impedance is
> changing drastically at the moment the diode turns on and current begins
> to flow. If the LO is a sine wave, that transition period is often
> considered to be about 20-30 degrees of the LO sine wave, and even longer
> if your LO drive is less than recommended. But by slamming a square wave
> into the mixer as the LO, you are turning those diodes on very quickly,
> (a few nS, which is a few degrees for the low LO frequencies we tend to
> use), which reduces the overall noise generated by the mixer.
The "noise" that you refer to; is it 1/f noise or some other noise source?
I usually see that the noise figure of a mixer is essentially identical
to its loss.
I am most concerned with the 1/f noise that Rick Campbell has noted in his
analysis of DC receivers which are of interest to me. If square wave
drive helps this, I would love to know. I know that square wave drive
improves the IP3 numbers.
I personally drive diode mixers from 74AC04 inverters. I parallel two
inverters and source terminate with a 43 ohm resistor. Following this is
usually a attenuator to drop the drive level to +7dbm for a SBL-1 or equiv.
The output impedance of the inverters is roughly the DC output resistance
of the device. (Got this fact from a TI apps engineer) This value is easially
determined with a variable resistor and power supply. I have found the AC
series of parts to typically have an output impedance of about 14 ohms in
either high or low state. Thus two in parallel gives an output impedance of
7 ohms with the remainder made up with the 43 ohm resistor.
As for the output rise time, two of the inverters together into 50 ohms
looks like about 2ns. Darn fast transsitions. The power consumption is
much less than a class A amplifier too. Also it is very easy to accurately
determine the drive level to the mixer. You can also make it variable just
by adjusting the AC04's Vdd line. This may vary the output impedance a
little bit however.
For the curious, don't try this with a 74HC04. The output impedance is much higher.
It is refreshing to see techinical discussions here. Thanks.
Search QRP-L Archives
QRP-L Archive |
[ 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 ]
This archive was generated by hypermail 2b29 on Thu Feb 07 2002 - 08:03:45 CST