Help Troubleshoot 2 FET amp.

Up to 30MHz. Mostly AM band where you have many, many signals that can cause mixing products. Mikek

Ought to be good setting next to the radio station! :-) I don't know how it it measured, although I know the author has several references to 6MHz and 9Mhz, looking for 15MHz products. I have seen at least one of his setups to measure, but couldn't recite it. He also labels it IIP, I don't know if that is different. I usually see IP elsewhere. I just know having it high is important. I'm following the directions to build a directional antenna for the AM band. Don't care to change anything until I have a working system. Mikek

Intermod is measured in dB, not dBm.

Not according to this PDF. photos.imageevent.com/qdf_files/technicalgoodies/.../Measuring%20IP1.docx.pdf

Or this video Video.

5:54

Amplifiers with strong feedback are quite linear for low signal levels and extrapolating the IP3 may give huge IP3 values. But at signal levels close to clipping, the IP3 values collapse rapidly.

For non-feedback amplifier, the computational IP3 point is typically

10-20 dB above the -1 dB compression point (which is close to maximum amplitude).

IP3 figures are of interest, since the intermodulation product can fall close to the intended frequencies and can no longer be removed with filters.

IP2 is seldom critical, except for the second harmonic of a single offending tone. If a receiver contains switching diodes for switching in front end filters, a very strong out of band signal can cause second harmonics in the switching diodes, not so much an amplifier issue.

The IP2 value is ridiculous.

When IP is used instead of IIP (Input-IP) or OIP (Output-IP) one should assume that the larger value is used at least for marketing. OIP = IIP + Gain (in dB), so an amplifier with positive gain, take the OIP value and label it as IP. For a device with negative gain (such as a diode ring mixers), the IIP is larger and is suitable for advertisement :-).

The IP3=41 dBm sounds reasonable for OIP3, since +41 dBm is about 25 Vrms into 50 ohms. Assuming -1 dBm compression point is 20 dB below this IP3 figure would be 2.5 V, which could be reached with a +9 V supply.

The band noise at those frequencies into a full size antenna is so high that it is very questionable if any preamplifiers are needed at all, perhaps attenuator would be more appropriate.

For very short whip antennas the antenna attenuates the signal all by itself.

Intercept points and compression points are power levels, which are expressed in dBm i.e. dB over 1 mW.

Take a look at ADA4817 maybe. 1 GHz jfet opamp.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

This amp shows numbers over 100dbm for IP3 with -15dbm input.

Graph shown near the very end of the pdf. I don't know the answer, but, it would be nice to find out where the disagreement stems from. Is there just an exaggeration (I doubt that), it is in the way something is interpreted, or something else.

Yes, it is stated that you can't use this on a full size antenna. One antenna is a 3 ft whip, another is a Flag or Delta with a ~900ohm termination, and an output less than -40dbi, a third is a 90ft BOG.*

• used in a quad phased array

With -15 dBm test signal and more than 100 dB range, the 2nd harmonic would be below -115 dBm. Depending on the spectrum analyzer bandwidth (more than a few kHz) the 2nd order distortion would be swamped by SA noise.

BTW, you are using a completely different amplifier topology compared to your two transistor (cascade?) FET stage and refer to single ended bipolar feedback stage. A similar bipolar feedback stage using

2xMRF586 in push-pull configuration at 12 V and 2x30 mA in is the N6NWP design with +48 dB, input-IP3.

You would need a source follower to mach the highly reactive whip to a coaxial cable.

Are you referring to a flagpole or delta loop ? The -40 dBi sounds strange, it is nearly as bad as a ferrite antenna.

Is the BOG a tree or what ?

With some loading coils you can tune out the capacitive reactance and feed a coax. This will give sufficient power without any amplifiers.

So, are you saying the 100dbm number is not true? Also. that was IP3 not IP2.

Yes, I only used that as an example of another very high intercept number, and also using dbm.

I still have all three issues of QST that have that receiver design article. (and corrections)

The output impedance of the 2 FET amp is listed as ~150 ohms. The author has given up on coax, because of signal ingress. Many of his systems are phased arrays and ingress degrades the pattern. He uses twisted speaker wire with success. and does not match the 150 ohm amp output to the 50 ohm receiver input.

Yes loop, the output is what it is. Look it up. By using the high input impedance 2 FET amp, the antenna is not loaded so has it's full output voltage. That's a 6db voltage gain. If you match the 900ohm impedance to 50 ohms, you would have 25db voltage drop. Flag or Delta have a single lobed pattern vs double lobe for a ferrite rod.

Beverage on ground.

Loading coils on what?

I have an active antenna that works from 200kHz to 30MHz, without any loading coils. Just a high input impedance amp and about a 1.5" x 1.5 inch piece of pcb as the antenna. It is comparable to my 135ft long wire antenna, except, I note the long wire antenna is slightly directional at some frequencies. After I get the second iteration of the 2 FET amp working, I'm plan to set it up at an active antenna with 3ft probe to see how it compares to this,

Mikek