Electronics > RF, Microwave, Ham Radio

How could I quantify a maximum allowable input for my home-brew RF amplifier?

(1/1)

MisterHeadache:
First, to be above board about who I am, I host a YouTube channel called Level UP EE Lab.  I used the ‘MisterHeadache’ username here before I started that channel, and I cannot edit my profile here to update my username.  So that’s why the discrepancy.

Anyway, one of my biggest projects to-date on my channel is a home-brew HF transmitter, and I am a bit stuck on the RF preamp that sits between the final mixer and the RF power amp.  The preamp works just fine, but I need to quantify its limits.  Specifically, I need to define the maximum input SSB RF power that I can feed it that (1) does not cause ‘excessive’ intermodulation products to be generated and end up as bandwidth increasing splatter and (2) stays below the maximum heat dissipation of the transistors.

Of course, the answer to item #1 has to be less than item #2, otherwise I have a really bad design.  And I do already know the ballpark for item #2 because during my initial trials to push it hard I blew the transistors and released the magic smoke :palm:.  But fortunately they are through-hole parts and are easily replaced.

FCC 47 97.302(b) offers some guidance about emissions resulting from modulation, but doesn’t quantify limits like parts (d) and (e) do for spurious emissions.  Rather, part (b) just states that such emissions "must not cause splatter or keyclick interference to operations on adjacent frequencies.”

So my questions:
1.   What determines ‘excessive’ when it comes to intermod products?
2.   Given an answer to the first question, how can I best determine what power limit meets that target, given my humble arsenal of test equipment (see below)?

I’ve read the ARRL test procedures manual and understand test 4.5 for measuring IMD products. However, my home-built SA doesn’t have a resolution bandwidth anywhere near the 100Hz required, but even if I could accurately measure the IMD products, so what?  It’s back to my question about what is excessive.

Limiting the input power is easy enough, I’ve got adjustable gain in the microphone amp that feeds the rest of the IF and mixer chains.  And I can monitor and report accurately enough the real-time transmit power, so there’s feedback available to the user when the gain is set too high.  Just need to know how to determine how much is too much.

And here’s the relevant test equipment I have: homebuilt SA mentioned above, Siglent SDS2202X-E (has FFT but I doubt it has anywhere near the capability needed for IMD measurement), two-tone signal generator (~700Hz and ~1900Hz at equal amplitude), Airspy HF+ SDR (which I can use to easily see the IMD products with appropriate attenuation).

Thanks in advance for any advice.  I’ll be sure to credit on the next episode.

DimitriP:
The dreaded IMD figures!
Summurising your post.
A) You want to know minimums/acceptable
B) Max input before IMD shoots outside the limits AND doesn't fry the devices
C) Can you trust your measurements with what you got

A) -40 -50 -60
B)  Once you are set-up for a you increase power until IMD goes nuts or the devices  reach max temp whichever one comes first.
C) Can't tell without knowing how well you homebuilt SA is. Easy guess is that might be time for an upgrade :)

many fun threads regarding IMD on Eham.
Here is one to start you off :  https://www.eham.net/forum/view?id=index.php/topic,135937.msg1259334.html%3f#msg1259334

Fun times!

MisterHeadache:
DimitriP, thanks for the reply! Wow, that thread has had a lot of activity. I started reading it, got to the tenth page, will finish it later today. But if I understand your suggestion, which aligns with that thread, using -40dB below primary for IMD3, -50 for the 5th, -60 for the 7th would be defendable practice.

Regarding my home brew SA, it’s pretty much the design Wes Hayward pubished years ago. Its got acceptable log accuracy (it uses the AD8307), but I’ve only got 300kHz and 30kHz resolution bandwidths, so wowefully too wide for this. I’m eyeing the TinySA ultra, its got a 200Hz resolution bandwidth but am concerned how long a scan at that resolution takes.

Navigation

[0] Message Index