EMRFD Message Archive 756

Message Date From Subject
756 2007-05-11 16:35:04 Kevin Purcell G4COL Experiments with high current Gilbert cell mixers
I've uploaded a copy of G4COL "Experiments with Gilbert cell mixers"
from Sprat 95 to the files area. I sent G4COL some email but I
haven't get a response.


He was disappointed when the SL6440 was discontinued and tried some
experiments measuring the IIP3 with "high currents" in a couple of
Gilbert cell mixers.

In the CA3046 transistor array he found with a total standing current
of 17mA he could get IIP3 of more than +20dBm.

With the 1496 set for a 11mA per side (22mA standing current in the
mixer -- a total current of 33mA including bias current) he found an
IIP3 of +20dBm.

A very nice result. Much better than a 602 or even a low current 1496.

Note he exceeds the absolute maximum rating of 10mA. Just do the
maximum power calculation (unfortunately not specified for the
plastic package but it must be less than 500mW or so). A finger tip
test would be useful!

No reports of noise figure but you can get an estimate of that from
the Moto app note (at the regular 1mA current setting)

> Another advantage of the MC1496 product detector is its
> high sensitivity.

> For a 20 dB (S + N)/N ratio, demodulated audio output
> signal, a 9 MHz SSB input signal power of –101 dBm is
> required.

That's an MDS of -121dBm. Assuming a bandwidth of 3000Hz then gives a
NF of 18dB for the 1496. Which is not certainly great but not bad
either below 30MHz. For example, this would only be above the
atmospheric noise figure in a "very quiet rural" on 10m (assuming a
loss of 3dB in bandpass filters). But there's not much margin there.

It could be a useful device for designs that you expect to build
around the world (it's cheap and easy to get).
Kevin Purcell
773 2007-05-20 13:43:56 Kevin Purcell Re: G4COL Experiments with high current Gilbert cell mixers
If you are playing with a high current 1496 Peter Chadwick, G3RZP,
gives a useful rule of thumb for calculating the chip temp from the
power dissipation.

The maximum dissaption for the plastic DIP (NJM1496) is 570mW (and a
lot smaller 300mW for the surface mount package).

The thermal resistance for the plastic DIP is 100 deg C/W. This is a
lot less than the 220 degC/watt for the (now unavailable) metal can

From the app note you can make a more accurate estimate

> Power Dissipation
> Power dissipation, PD, within the integrated circuit package should
> be calculated as the summation of the voltage−current products at
> each port, i.e. assuming V12 = V6, I5 = I6 = I12 and ignoring base
> current,
> PD = 2 * I5 * (V6 − V14) + I5 * (V5 − V14)
> where subscripts refer to pin numbers.

Or as V14 is usually ground

PD = 2 * I5 * V6 + I5 * V5

The (ballpark) total power is 2 * I5 * Vcc (ish). With 14 volt supply
and the 10mA "max" programmed current you won't exceed the maximum
dissipation though the chip will run hot at 50C or so (just too hot
to touch) but within spec. So you might push the current up a bit
beyond the "max" though I suspect the wire bonding the chip to the
pin is the limit.

> RE: MC1496 use on 6Mtrs?
> by G3RZP on January 17, 2006
> You can improve the carrier suppression with some external trimming
> across the signal input port. Use a 50k pot witn the slider to the
> negative supply line.
> It's possible too, to push the current up a bit with external
> resistors. This improves the signal handling. Just make sure that
> the total current times supply volts doesn't give enough power to
> get it too hot. Take the watts input and multiply by the thermal
> resistance junction to ambient: this tells you how much above
> ambient the chip temperature is. Add the ambient temp, and keep
> chip temp below +125 deg C. I can't remember the chip to ambient
> thermal resistance for the dual in line package: for the TO (metal
> can) device is about 220 degC/watt.