Need to Measure Transmitter Spurs, w/o Spectrum Analyzer

Discussion in 'General Technical Questions and Answers' started by KK5JY, Oct 19, 2017.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Left-3
ad: abrind-2
ad: L-MFJ
ad: Subscribe
  1. K6BRN

    K6BRN XML Subscriber QRZ Page

    To do this (measure how many db below the carrier...) you will need a calibrated spectrum analyzer and an operator who knows bow to use it without blowing it up. There are other, more complex and demanding ways... which is why a spectrum analyzer is preferred.

    Aim for about 1 mw of input power to the analyzer, or 50db of attenuation from 100W or 30 db of attenuation from 1 watt. Your USB dongle likely requires less than 10 microwatts in order not to self generate a lot of spurs due to saturation somewhere in the signal path. Thats 50 to 70 db of attenuation. Is your unspecified attenuator that big (probably not). Most spectrum analyzers will be damaged by signals larger than +20 to +30 dbm (100mw to 1watt), so its REALLY important to have a good, stable attenuator of a known value.

    Brian - K6BRN
     
  2. K6BRN

    K6BRN XML Subscriber QRZ Page

    BTW - you will not be able to build "..an 80 db resistive (RF) attenuator..." that does not leak like a sieve compared to its attenuation value. Sorry, the impedance matching and shielding is not something YOU will be able to do. There are plenty of 10, 20, 40 db fixed power industrial attenuators on ebay that can help, but be prepared to make an investment. And variable attenuators, like the many HP units available used, are incredibly useful for this work. You are WAYYYYY over your head.

    Your best bet is to find someone who already knows how to do this and has the equipment, then have that person tutor you in how to do it and why to do it that way. This is not an insult, it's an excercise in knowing what you do not know and seeking to learn.

    Good luck.

    Brian - K6BRN
     
  3. KK5JY

    KK5JY Ham Member QRZ Page

    I understand where you are coming from, and if I had ready access to that kind of gear, I would already be doing it. The reason I asked the question originally was to try to find a way to do this without thousands of $$$ worth of test gear. Hams have built their own gear for decades, and it seems to me that if they are going to continue to do so, there need to be techniques that can allow them to do so without needing access to an LMR-shop's-worth of diagnostic equipment.

    I did some searching on spectrum scopes, and I see that they are a lot less expensive than they used to be. I can get a "headless" scope (needs a computer for the display) for a couple of grand. So it's not like I need to lay down $50k or $100k on a big HP box the way my dad's generation did, but even even at $2k, the test gear cost exceeds the project cost by nearly two orders of magnitude. That's not really practical for a homebrew project budget.

    That's also not a complaint against your post. :) It's just that having the ability to experiment isn't worth very much if doing so means that I need so much test gear that I could have just gone out and bought a contest-grade station for the same amount of cash. It seems that there should be a way for the average Joe Ham to build his own gear, and know that it is reasonably compliant with the emissions standards, without having to find/borrow/buy lab-grade test equipment.
     
  4. K6BRN

    K6BRN XML Subscriber QRZ Page

    Matt:

    Did you really look, because excellent spectrum analyzers can be had, brand new for less than $1300, or much less than that, used on ebay, or even less for pretty poor kits that may work on lower HF frequencies. BTW - you'll also need a selection of attenuators and filters to do this well - ones that are broad-banded, and very RF tight. More $$$

    But - I understand what you are saying. AMATEUR-grade measurements should not be EXPENSIVE.

    The problem is you are mixing a compliance level measurement that requires precision, accuracy and relatively high dynamic range with a home-brew tradition that usually gives only "ballpark" results. Life and physics is rather unforgiving when all three are demanded at once, and amateur kludges will not provide precise and accurate results.

    So, if you want "ballpark" results, simply transmit a CW carrier into a dummy load while any receiver with an S-meter is used to pick up the main signal leakage, using a small wire stub antenna. Then scan the receiver +/- 200 KHz around the main carrier looking for "birdies" that come and go with the keying of your homebrew transmitter. Note the signal strength of each birdie and make a table of them in signal strength vs. frequency. Assume at each "S-Unit" is 6db and calculate the difference in db, for each birdie, below the main carrier. This method should be obvious to you. And it WILL reveal some major issues, if they exist, but not more subtle ones. Hams have been using this approach for many decades.

    The problems with this method are accuracy and dynamic range. Every receiver will give a slightly different result, because most S-meters are not calibrated, and the most range you will likely be able to get, on a very good day, is about 35-40db or so before the reading becomes totally meaningless due to s-meter and ALC non-linearity (because S-meter readings are usually just receiver ALC voltage) And here are many other issues with this approach. On the other hand, there ARE high accuracy, calibrated receivers that can be for this purpose. Problem is, they are also in Spectrum Analyzer cost territory.

    So is a ballpark test good enough for your purposes, or do you really want to measure "compliance". Your choice. One costs, the other, not so much.

    Brian - K6BRN
     
  5. KK5JY

    KK5JY Ham Member QRZ Page

    Yes, this is the core of my original question to the group. Hams 50y ago didn't have spectrum scopes, so they had to find a way to make their home-brew gear compliant, even with the lesser restrictions that were in place at the time.

    The measurement doesn't have to be accurate to a single dB. It only has to be accurate enough so that even with the margin of error, there is a high confidence that the transmitter is still compliant. E.g., a margin of error of +/- 15dB is fine, as long as the strongest spurious signal measured is around -60dBc.

    I have to admit that I am a bit surprised that this question doesn't get asked more often.
     
  6. G0HZU

    G0HZU QRZ Member

    ...or you can use a clean sig gen and a decent DB mixer and a PC with a decent soundcard as previously suggested. This will be able to measure relative signal levels with log detection performance approaching lab grade gear over a range of about 60dB or so. The major downside is that it will be tedious to manually sweep the sig gen over the whole signal range whilst looking at the soundcard.
    This could easily take several minutes per sweep. But it should be able to measure harmonic distortion and close in splatter/clicks with similar performance to a decent old school lab spectrum analyser. I'd expect it to totally outclass the classic $$$ HP lab spectrum analysers that were made >25 years ago for the modulation tests. But you do need a decent sig gen for the LO that can be tuned easily and quickly. I'd expect that most old (non synthesised) lab sig gens from the 1960s - 1980s would be OK here and they are usually quite cheap to buy nowadays. A decent DDS could be used here for the LO although an AD9850 would be a bit marginal due to the carpet of spurious it produces. But if you only care about -40dBc then even an AD9850 DDS could be used for the LO across the HF band. But it won't be a rewarding experience to see just how dirty these old AD9850 DDS chips are. I first played with the AD9850 DDS just over 20 years ago!
     
    Last edited: Oct 21, 2017
  7. G0HZU

    G0HZU QRZ Member

    I think this forum is often kind of stuck in the 1960s-198os. Threads like this often get dominated by ageing 'experts' that tell you that everything today is as difficult (and expensive) as it was for them 30-40 years ago.
     
    Last edited: Oct 21, 2017
  8. KK5JY

    KK5JY Ham Member QRZ Page

    I have to admit, I really like your idea. The reason I'm hesitating on it is that I'm not sure I have the RF construction skills (or components, for that matter) to make it work reliably.

    Just for this project alone, I have built and tried several different RF amplifier/driver circuits, and none of them have the real-world soldered-together performance that the data sheets promise. I'm pretty sure the problem isn't the individual parts, per se, or even the overall construction, it's just that the in-circuit reality is more complicated than what the manufacturers did on a test-stand.

    I'm keeping your idea in my queue, and I may end up having to go that direction, but if I have to build an active RF circuit from scratch, in order to measure another from-scratch RF circuit, I'm just adding to the uncertainty of the result? Perhaps I am missing something about your idea that would make it more fool-proof than what I understand?
     
  9. KA9JLM

    KA9JLM Ham Member QRZ Page

    Why make it so complicated ?

    Take it to the FCC and have it certified.

    They have the proper test gear. o_O
     
  10. KK5JY

    KK5JY Ham Member QRZ Page

    Your helpfulness is ... underwhelming. ;)
     
    KA9JLM likes this.

Share This Page