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: abrind-2
ad: Left-3
ad: Subscribe
ad: L-MFJ
ad: Left-2
  1. G0HZU

    G0HZU QRZ Member

    Obviously, the best solution would be to borrow or buy a decent spectrum analyser. The mixer/soundcard method I suggested is a good one in terms of performance but it will rapidly become tiresome if you wanted to do lots of tweak/retest work on your transmitter. This is because each manual sweep will take several minutes compared to a second or less for a decent spectrum analyser. However, even if you get access to a proper analyser it's worth making the mixer/soundcard setup because it will totally outperform all but the very best modern analysers if you also want to look at the close in spectrum of OOK waveforms at slow sending rates. The PC/soundcard will be very fast and fluid here compared to a traditional swept spectrum analyser :)
  2. KK5JY

    KK5JY Ham Member QRZ Page

    This is a good point, but if the way to solve my original problem is to build something, I expect I will have much better luck building a strip of attenuators than some other kind of circuit. I'm only starting with maybe 6.5W on the lowest band, and much less than that on higher bands, so it shouldn't be too difficult to get the power down to a level where the receiver (whatever that is) can manage it.

    BTW, I have a couple of receiver options, including one of these:
    ... and one of these:

    The latter obviously has an advantage with respect to dynamic range, but I'm not sure that it is otherwise that much better than the former for this purpose.
  3. SM0AOM

    SM0AOM Ham Member QRZ Page

    This is a dilemma. Radio amateurs are indeed supposed to ensure that their transmitters are compliant to the spectrum limits set
    out in both the ITU Radio Regulations and in national rules.

    Most amateurs just trust the designer or manufacturer of a piece of gear, and never check up the actual performance. I have to admit designing and building a 144 MHz transverter in 1972 that I am quite sure that I would not have dared put on the air if a spectrum analyser had been available.

    For "ordinary" transmitters with the class-C output stage and the Pi-network usually provided a suppression of the 2nd and higher harmonics of better than -45 dBc when properly tuned. This was seldom sufficient to prevent TV-band 1 TVI, so an external low-pass filter was usually added.

    The few times amateurs actually bothered about harmonic suppression in the "pre-SA" days, they usually used a step attenuator and a receiver with an S-meter. I recall a former colleague manually measuring the spectral purity of an entirely home-brew 14 MHz SSB transmitter using a borrowed Hammarlund SP-600 receiver, a dummy load and a 0-100 dB step attenuator. He was able to identify products to about -70 dBc.

    An interesting aspect is that most tubed linear amplifiers,both commercial and home-brew, that include the 50 MHz band do not meet neither ITU nor FCC harmonic suppression requirements at frequencies above 30 MHz because the pi-network ultimate rejection degrades with frequency when a separate LP-filter is not used.

    We traditionally have great liberties in home-brewing and modifying equipment, that frankly go far beyond the actual technical knowledge shown by most
    "instant hams" and appliance operators of today. Operators in other radio services that show comparable technical knowledge are not allowed to home-brew and modify equipment. Radio amateurs are supposed to form an elite of radio hobbyists.

    In my opinion, it is only a matter of time before the authorities will remove our home-brew and modification privileges due to lacking technical knowledge.
    A stop-gap measure would be to have an engineering degree as a requirement for the highest licence class, and reducing the privileges for the lower classes accordingly.

    Last edited: Oct 21, 2017
    KA9JLM likes this.
  4. K6BRN

    K6BRN XML Subscriber QRZ Page


    Eh... Sonny? Say again? Respecht yer elders, I say. :)

    Ummmm... back to reality....

    Nope - your approach is an uncalibrated, overcomplicated Rube Goldberg approximation with so many uncontrolled/uncalibrated variables that it falls in the "Ballpark" category I outlined above. Does not mean it won't work to detect gross problems. But you will not be doing any "certifications" with it.

    A spectrum analyzer test can be a two minute affair - its really a fast, simple test, if you know how to do it. Bob Allison of the ARRL was running quick tests outside of the exhibits at Hamcon West in September, using a relatively low cost (but calibrated) Rigol DSA815, for anybody who dropped by with an HT, and he was even issuing certificates of compliance. Total test time, about 15 seconds per run. So much for being complex. And the DSA705 is cheaper still and would have worked as well. But Bob knows what he's doing, to begin with.

    Worst of all, you didn't ASK the RIGHT question: "Does the O.P. really need to PROVE compliance through testing?" Answer: NO

    Second RIGHT question: "Does the O.P. need to detect gross problems with the homebrew transmitter?" YES Which means the O.P. can probably get away with any of many ballpark checks for undesired emissions and does not and never has needed a spectrum analyzer. But he should do some calculations regarding output filtering and possible spur generation to at least show his creation is UNLIKELY to be problematic.

    Ask anybody who has built a homebrew amplifier or modified an amp about this. Not too many run compliance tests. But they do tend to use good design practices and run a few simple calculations.

    BTW - I'm probably younger than you. But old enough to know how to make things simple and accurate - when needed.

    Also, just sampled a few choice single-malt whiskies over here in Edinburgh, so I may be just a little less tactful than usual. Sorry.

    Brian - K6BRN
  5. KK5JY

    KK5JY Ham Member QRZ Page

    This is an interesting point. One thing I was hoping to do was to drive a tube/valve linear amplifier using this transmitter, and since the output has a Pi-L network, I had hoped that it would guarantee that the harmonics were more under control at 800W than the transmitter alone at 5W. Perhaps that's not so? Or were you specifically addressing amplifiers that included 6m?

    I'm not sure what the LPF is that I'm using. It's three series inductors with four shunt capacitors. I don't know what one would call that... Pi-Pi-Pi?? It seemed to have good performance in the measurements that the kit designers did.

    It's true that most hams wouldn't even know where to start to do something like this. Even the engineers. I have a background that includes quite a bit of LMR, so at least I have exposure to the issues that need to be overcome. I also took the extra step of getting my commercial licenses, but to be honest, studying for those tickets didn't impart any more knowledge than a US Extra class license, even with respect to this kind of subject. It's something to hang on the wall, but the General Radiotelephone license that my dad had back in the day was much more intensive with respect to radio design and maintenance. My General Phone license is an amateur-hour accomplishment compared to his. ;)

    You may be right that we may someday lose the ability to build gear. But it would seem an unnecessary step, since people have been doing this for decades. Then again, the -30dBc requirement that the FCC used to apply to this kind of project is a very different target to hit than their current -43dBc target. So maybe the world today is just different than it was when everybody home-built their own gear? If the target was still -30dBc, I would probably not feel the need to ask you all for your wisdom. :cool:
  6. G0HZU

    G0HZU QRZ Member

    What is it with radio hams and hobbyists and the need for 'calibration' and 'certification'?
    He's just measuring spurious from a dirty DDS based homebrew transmitter for goodness sake.

    The method I suggested is for making relative measurements in terms of dBc. Nobody in authority (with any sense) really cares how accurate the power level of the 5W carrier is. For making relative measurements down to about -50dBc, the log performance and linearity of a decent mixer and soundcard will be about as good as the flagship spectrum analysers from HP that used to cost $85k 20-25 years ago regardless of how recently that analyser was calibrated or certified. The conversion efficiency vs frequency of a decent wideband DBM will be fine for stuff like this.
  7. K6BRN

    K6BRN XML Subscriber QRZ Page

    Karl - Arne:

    While I largely agree with you, amateur radio was explicitly set up as a relatively safe "crib" for people of mixed skills with power and spectral limits that largely contain any real damage an errant amateur can do, as long as they TRY to play wihin the rules. So I don't think the FCC will be cracking down on home-brew equipment builders very soon. They barely have the funding to police their current scope, with regards to amateur radio. The simplest thing do do would be to close out amateur radio entirely.

    Which is why I very much appreciate ARRL's lobbying work to keep our bands open. It's easy to criticize ARRL for its imperfections - but they'd be VERY hard to replace, in practice.

    Bria - K6BRN
  8. KK5JY

    KK5JY Ham Member QRZ Page

  9. K6BRN

    K6BRN XML Subscriber QRZ Page


    Ahhhh... no... Relative or not, a sound card based measurement will not be about as good as any purpose built professional equipment from the last two decades and nobody will accept it for an accurate measurement. But it really does not need to be, as I said above.

    You've never worked in field, have you?

    Regarding "certification", this relates to the O.P.'s original desire TO BE SURE.... etc, etc, etc...., which, if you finished reading my reply to you above, I said he did not need to close hermetically.

    Patience is a virtue that can come with age, grasshopper. One must read carefully before framing a hasty reply.

    Brian - K6BRN
  10. KK5JY

    KK5JY Ham Member QRZ Page

    By the way, do you have (or have you seen) a schematic for a design that uses the DBM you suggested, the way you suggested?

Share This Page