PA1DSP
   
Articles
Mixers
Signal Processing Concepts
Complex numbers (part 1)
Complex numbers (part 2)
The RLC network
Why NOT to use the NE602

Why NOT to use the NE602

Many simple receivers I see around the web are based on the NE602 (a.k.a the SA602, NE612 and SA612) chip. This chip packs a mixer and oscillator into a tiny 8-pin DIP package. It is cheap and easy to obtain. The (fully differential) mixer offers some 18 dB gain. The local oscillator drive-level is low, which saves batteries in portable equipment. No wonder people use it a lot!

Unfortunately, the NE602 suffers from one major (and disastrous) drawback: It has an extremely bad large-signal characteristic. The datasheet tells us that its third-order intercept point (IP3) is around -13 dBm.

It also shows the following compression graph:

The desired output of the mixer (a.k.a. the fundamental mixing product) starts to compress at around -25 dBm. This can be seen in the graph as the ‘knee’ in de the fundamental product line. The knee shows that increasing the input level by an amount no longer produces an increase at the output of the same amount. The mixer is no longer linear and starts to produce harmonic distortion. Increasing the input level further will result in spurious signals in the audio and desensitization of the receiver.

Let’s see how much signal would be needed at the antenna of a receiver to reach -25 dBm at the input of the mixer. We start the analysis by converting dBm’s into milliwatts:

The input impedance of the RF port of the mixer in the NE602 is around 1.5 kOhms. Therefore, the peak input voltage at which distortion starts will be:

So when the AC swing at the input of the mixer exceeds 0.1 V peak, distortion will occur.

A well-known 40 meter receiver based on the NE602 uses this input network:

The input network has the following frequency characteristic (strange how it peaks at 6.4 MHz, which is not the 40m band!):

The mixer input reaches 100 mV at 6.4 MHz when the antenna voltage is 0.04 Vpeak. In this case, the power delivered by the antenna is 13.2 microwatts.

It doesn’t take much to generate such a signal at the antenna; a nearby AM broadcasting station will easily cause distortion or blocking of the NE602’s mixer. In fact, almost anything nearby will achieve this!

Given the above, it is clear that strong signals must be kept away from the NE602. A very good preselector together with good shielding will help in reducing the chances of distortion and desensitization. However, unless you live in a secluded area far away from everything, I recommend against using the NE602 as a front-end mixer.

On the positive side, the NE602 can be put to good use in a transmitter, as the builder is in complete control of the signal levels. It is also acceptable as a second mixer (after a good crystal filter), which is its intended use.

Concluding remarks

In my opinion, any (simple) receiver that uses an NE602 as its front-end mixer must be considered a toy. Incidentally, the K1 from Elecraft uses an NE602 chip as its first mixer. The designers reduced the chances of overloading the mixer by putting a 9th-order filter at its input. Even with this amount of filtering, a selectable 14 dB attenuator was included to cope with high-level in-band input signals.

We can still build very simple receivers with the NE602. We can use ready-made diode mixers, such as the SBL-1 or the TUF-1 from Mini-Circuits. They cost quite a bit more, but I think they’re worth it. Or maybe you should buy them “because you’re worth it!”. The downside of these mixers is that they require a high oscillator drive level (at least +7 dBm) to work and they don’t include an oscillator transistor like the NE602 does. They must also be terminated properly at every port, which complicates things a little.

Another alternative to the NE602 is the 74HC4066 analog multiplexer. This IC can be used as a mixer on the lower amateur bands and it doesn’t require a high oscillator drive level to work. It has been successfully used in a 40m rig by Oliver DL7JGR. Additionally, Sprat number 99 features a 74HC4066 based mixer by Leon VK2DOB.

Whatever your next receiver project will be: Please, avoid the NE602 in receivers (unless you include lots of RF filtering). For transmitters, the NE602 is great!

73,

Niels.

_______________________________________________________

 
 

Do you want to know more about IP3 or how to ‘plan’ your next receiver? I recommend ‘Experimental Methods in RF Design’ by Wes Hayward et al.

This book is the modern version of the well-known ‘Solid State Design for the Radio Amateur’. It features many circuits covering amplifiers, oscillators, mixers and more. It comes with a CD packed with additional articles and software.

It is a great buy; no experimenter should be without it!