The AM Forum

Has anyone else noticed that there seems to be more and more transmitters operating AM that have been spitin' and sputin' up and down the band?  The audio of these stations may sound fantastic on frequency, but give a listen up and down the band a little and see what you think.

The Mighty Elmac net on Wednesday nights on 3.880 has been getting hammered lately.  5 KC separation will cause adjacent channel interference to some degree, but it's much worse than that as I can hear it at least +/- 10 KC on some of the signals.  I'm sure it's not intentional, but it's always a good idea keep an eye on your transmission characteristics.

In a recent topic on the "AM Window", K4HX posted Part 97.101.  I enjoyed reading this part:

"Emissions outside the necessary bandwidth must not cause splatter or keyclick interference to operations on adjacent frequencies".

I love operating the AM mode and with that comes the responsibility to do my best to keep a clean signal.

Pretty much two separate issues Dave, though I can see the relationship you're referring to.

There are indeed some nights when folks buckshot up and down the band. Sometimes it's related to really good band conditions, other times it's due to the amount of power and audio response used. I heard a couple 5 Landers on 3.885 complaining about a 1 Lander on 3.875 last week whose buckshot was clobbering them which was probably a combination of both. Certainly a valid issue.

The other issue is that of holding a net in a busy part of the band where you're bound to get interference locally or long haul when the band goes long. It might be a good time to rethink the frequency used, regardless of how long it's been on XYZ frequency. Especially if/when the frequencies around you are already occupied when you fire up. 5 KHz certainly is reasonable, but so is having a backup plan for when it's noisy.

If you're running the net and someone pops up nearby, certainly contact them and politely inform them of the situation. Most folks appreciate being informed if they're creating problems. If that fails or the frequency or those nearby are already in use, have a backup plan: pick an area down the band a ways and give folks a +/- idea of where to look. Maybe change to a later time when things thin out a bit. 

Or you can continue on as-is, accepting the noise and interference common in that part of the band most every night during prime time. It's one of those 'six of one, half dozen of the other' situations.

Good points Todd.  Yes, it's good to be flexible with our frequencies we are using and turn that VFO dial when it will help.

It's important we do our job to be good neighbors.  Part of this is to make sure we are doing our part by keeping our transmissions as clean as possible.  Having AM QSO's within 5 KC during prime times is definitely doable.  In fact, prime time is when we should be particularly diligent with good operating practice.

Have you ever noticed what a touchy situation it is to suggest to someone that they are causing interference up and down the band?  I find it difficult.  It's also difficult to graciously except criticism.  Not to mention, that some might perceive that an AM station is causing problems and it may well be the listeners receiver, passband, etc.  All part of the game.  Good people skills definitely helps on both ends!

I enjoy reading your well thought out responses on the AM Forum Todd.  It's a great place to communicate.

Dave

I disagree that a 5 kHz spacing between QSOs in the same geographic region is good practice. Even if all parties have very clean transmitters, run 3 kHz audio and use 6 kHz receive bandwidth, there is a 1 kHz overlap. In other words, interference, and probably a large amount of it will occur. Add in the realities of less than perfectly clean transmitter and not having a brickwall cutoff at 3 kHz on the audio and even more interference will occur.

Should all strive to run a clean transmitter and limit their high end audio when the band is crowded? Definitely. But 5 kHz is too close.

Hi Dave,

Yes, for the most part, I think most folks appreciate being informed of a LEGIT splatter problem.  However, it sometimes seems that the worse the problem is, the worse defensive attitude some have.  It's like they have gotten so beaten up about it, they go into denial. There's a couple of stations on ssb that hang down around the DX window that have such severe QRO splatter that guys have given up trying to help them. These two get belligerent and scare anyone off that tries to help.  It's actually funny, just like approaching a nest of yellow jackets.  Huzman and I have joked about it for over a year now.

But I think most guys are actually shocked to hear they have a problem and will do everything possible to correct it - even QRT on the spot and do some tests into the dummy load.


I feel anyone running a big AM rig has a responsibility to have "dynamic" control of their rig . ie, a way to limit audio bandwidth and audio power based on what is going on around them.  Many guys have audio filters in the form of EQ's, SDR radios, brick wall audio filters, etc to give a tight 3.5 - 4.0 kc response if need be. Backing the % of modulation down helps tremendously too.

But the most important thing is to have a CLEAN transmitter in the first place - with good IMD numbers.  I've found the easiest way to test a rig is by using simple voice programing and watching an SDR spec analyzer (a simple SoftRock is OK) to see just how clean the transmitter is. A few hours of experimenting into a dummy load with all kinds of modulation levels and audio filters will tell you just how close you can get to another QSO, within reason.  We want to KNOW with certainty how clean or dirty our rig is without having to axe our buddy on the air.  Most of our friends who take a quick listen (without an SDR or spec) will say we are clean - to be nice and give us the benefit of the doubt - but do it right and run the full tests.

I remember a guy once down on ssb calling CQ in the DX window. He was as clean as a whistle. But someone broke in playing games and said he was splattering all over the band.  This guy came back and said, "No I'm not - I'm watching a spectrum analyzer as I speak" -  and then laughed and continued on. He was right and had the confidence to say so. I always remembered that and vowed to someday have that capability and confidence in my signal  too.  I finally added an SDR RX and can see my own signal now and know the limits I can push thangs.  

Just as for the last 90 years, there will always be transmitters on the air with problems. It will always be this way with the current technology and human control of the knobs.  One of the most difficult challenges in ham radio is to put together a station that runs high power and is really clean. I have been on both sides of the struggle and know how difficult it can be to pull off.

T

There's also absolutely no requirement to center the carrier in your receiver's passband, if you have a separate receiver/transmitter setup.  If you're in a QSO on 3880 and someone's on 3885, and you RXer has a 6 KHz bandwidth, tuning down to 3878 or so will keep your QSO partner's carrier in your passband AND greatly attenuate any adjacent interference.  It'll also give you a bit better audio response.

Too many hams think they need to tune in a signal for max S-meter reading, or keep the receiver right on 3.880.000.  It's a psychological thing, I guess.  But using a separate receiver (or a ricebox with dual VFOs) gives you great flexibility - ya just gotta use it.

Good point Johnny. This is also true if you have the right kind of synchronous detector.

Several good points! I overlooked one of the biggest issues you mentioned, Steve: geographic location. Being out of the hornet's nest on 75 since moving south, I don't notice the interference so much down this way when I got up to that part of the band. But it all came back to me when you mentioned that. I recall even discussing it with someone last fall at Deerfield. 5 KHz away from big station or group just isn't enough, hence the problems up there on a regular basis.

The passband issue I've come to take for granted since the 75A-4 makes it so easy. Tune in the signal, rock the passband tuning knob left or right depending on where the interference is coming from. Gone! 75 meters is not the place to use an old receiver lacking necessary noise-fighting features on a busy night.

Having a means of monitoring your own signal visually would be great. When Rob/AEX posted his videos from the HMR I was amazed at how clear everything became, being able to see just how wide I was or wasn't. Just because I don't bother anyone at a distance doesn't mean I don't have the possibility of doing so more locally. Maybe working a Softrock into the station would be a good idea. Just completed the 'JJ/'HUX scope sampler, so anything is possible!  ;D

I want to pick up on Tom's comment on using a SDR as a spectrum analyzer. Technology today is such that we all can be preemptive about our spectral purity.  Given RF sampling technology and SDR technology, spectrum analysis is within the grasp of many hams.  For those who do not have the resources, they can arrange to have a ham of known expertise to give a look at his signal and see how good it is.  The key is to be preemptive. 

Part of what sets us apart is that we are amateur radio licensees.  This really should place us on the same plane as the pros except we do not get paid for it and some may not have the same resources as the pros. But having said that, there are hams out there who would have the resources and would be happy to assist in helping any who wish to have adjacent channel purity checked. 

Again, we should want to be good neighbors by maintaining spectrally pure signals.  We have many more tools today to achieve this objective. Let's all resolve to take advantage of these tools.

Al

Has anyone else noticed that there seems to be more and more transmitters operating AM that have been spitin' and sputin' up and down the band?

Yes Dave, I have noticed this trend.

73
TM

Nearly a dozen years ago Steve, then known as WB3HUZ, wrote an excellent tutorial on the spacing of AM conversations. The book from which he began his research is now 50 years old¹.

http://amwindow.org/tech/htm/ambw.html

The technical points are worth keeping in mind. His conclusion, backed by what the book had documented, is relevant to the spacing of the Multi-Elmac Net at a time nearby signals are (I guess) perhaps 20 or more db stronger than those among net participants.


Not only is it more satisfying to maintain adequate spacing, it is especially important for "tall ship" stations to maintain a clean, well-defined modulation envelope. I believe this is a shared responsibility as we array our conversations along the band.

Thus Dave, your point is well taken about how and whether that is consistently achieved during your encounters. Those behind strong signals with high datarates have a particular need to monitor against overmodulation ("splatter") and whether their otherwise clean AF response might be better tailored during congested band conditions to stay friends with the neighbors.

---

¹"Handbook of Radio-Frequency Interference," volumes 1 to 4, published by Frederick Research Corporation, Wheaton, MD

Great point John and the major reason I really like the Hallicrafters SX-96 family of receivers (SX-100, 101, 115, 117, 111) with switch selectable sideband on AM.  These allow you to instantly choose the sideband with the least interference and you can operate on 3885 listening to AM USB without any interference from 3880.  Hammarlund HQ-170/180 receivers offer this ability although in my experience the alternate sideband suppression isn't as good as the electronically switched Hallicrafters system.  Other receivers which do a good job of this include the often maligned (and misused) Heathkit Mohawk and the  Drake 4 line, R-7, and R-8 receivers.  

The Central Electronics sideband slicer and the similar RME adapter use the phasing method on receive and do a pretty good job of alternate sideband rejection and these can be used with any receiver with a ~455 Khz. IF.  A TMC GPR-90 with GSB-1 provides exceptional alternate sideband suppression but even with careful tuning the audio really is tight with this setup and sounds better using exalted AM (locally supplied carrier) so that the actual station carrier doesn't have to fit within the narrow 2.5 Khz. bandpass of the 17 Khz. IF.  A Hammarlund HC-10 is the most practical of the external adapters and it provides a variety of bandwidths along with built in AGC and audio systems making it easy to add selectable sideband on AM to any receiver with a 455 Khz. IF output.b

The simple single crystal filter in Position 1 along with using the Phasing notch does wonders for those of us not blessed with fancy radios and detectors.

And a HRO-60/NC-183D  takes care of the bandwidth issue anyway but then y'all be whining they are too sharp.
If you want it both ways try 15M and above.

I don't know anyone who operates AM at 2.5+2.5 bandwidth so I'm not sure why anyone would run on 3880 and assume that 3875 and 3885 would still be options for others. Most people I see operating AM run 5+5 or even 6+6 bandwidth when they're in the window. Now and then I'll see someone running 12+12 and using the whole window for themselves, but that's a different topic. At any rate, up here in the northeast, a single QSO on 3880 really puts the squeeze on everyone. As stated several times already in the thread, when someone decides to pick 3880 to operate on, firing up on 3885 will obviously result in overlapping sidebands with 3880 and subsequent difficulty for those without selectable sideband capability. Sliding up higher than 3885 is impossible for many of us in New England due to the 24/7 blathering from the K1MAN eternal broadcasts on 3890. Setting up lower on 3875 results in the same sideband overlapping with 3880 and depending on the time of day, sliding down to 3873-3870 brings storms of protests from the daily activities of the 3864 sideband group or the boneheads who run sideband on 3868 as soon as they see activity lower than 3875. When 3870 - 3890 is wide open, I always avoid starting up on 3880 for the same reason that I don't take up two parking spaces at Walmart. It just seems logical to go high or low to make room for another QSO. At any rate, if I was going to run a weekly net and wanted a clear shot, I'd stay away from the middle of the only place many people choose to operate.

Rob W1AEX

We used to have the problem in the mornings quite frequently.  Historically, an New England group would be on 3885, and a Western NY group would be on 3880, and it was often a problem.  5kHz *is* in the normal audio spectrum....

We often have a group from the South and a group from NE on 3885 and 3880 respectively, but we are looking at diminishing conditions at that point, and not the other way around - so the interference is waning rather than increasing, as it is at night.

If you ever hear me on at night and the high end is excessive, I don't mind a reminder if it's causing interference.  I usually cut it down a LOT (with switchable filters) at night, but sometimes forget - which did happen on the Saturday of the A.M.T.R.  I noticed it, and the next night things were restricted to 4.7kHz at the highs with 14 poles.

Well said, Steve.  I think with the tremendous bandwidth of the class E  comes the responsibility of bandwidth management.  On nights / days where the spectrum allows, then the bandwidth can be increased but on nights of congestion then bandwidth should be decreased by the responsible station operator.  

On a different note, I'd like to encourage spectrum monitoring by the station operator so that the bandwidth can be adjusted preemptively.  I don't think things are changed by this bandwidth management as far as channel spacing is concerned (there, I've said it - "channel spacing").  I think we should consider what others have suggested and use 10 KHz as the norm for channel spacing rather than the exception.

Al

This is what I was thinking of for overall bandwidth as a reasonable average. Roughly speaking, 5KHz above or below the center frequency. If only things worked that simply on the average night.

The video that Rob posted from the HMR was an eye-opener for me. I've had Steve/HX and others check my signal numerous times, but seeing it with my own eyes was great. While it was within 10KHz overall, it was pretty clear that any minor change could have an impact. As you mention Al, always better to be proactive than reactive. A scope can give you a good idea what your outgoing audio is doing, but not where it's doing it. I know nothing about using a panadaptor for viewing my transmitted signal, but it's something new to learn. Tom's example of the fellow being able to dispute the hecklers is another good reason.

As you said Todd, it's never as simple in real life! Which station would you rather be operating 10 kc away from?

Rob W1AEX

It's also nice to operate on 3705 , 3715, and3733 as I saw a couple of nights ago on the qs1r.  Everyone was happy and the sigs were so magnificent that only one or two ssb qsos were interspersed. 

Nice to Get out of the self flagellating ghetto.

Of course we could always use those old phasing rigs and select the AM sideband and with the other side being down 40dB or better the 5KHz problem would go away.

The downside of using one SB Carrier is that I frequently will slide the IF shift on the side with the least amount of crud and with one SB that option would be removed.  Even with the FT301 I used to tune on the side of the filter that affords the best copy.  Again, with one SB that options goes away

Al

Or we could just use sideband. Or better yet, do not transmit at all - no interference.   :P

With the phone band expanded, there is no reason to jam AM QSOs cheek to jowl. If you choose to do this then you have no legitimate reason to complain about interference.

Dave,

I find most AM stations want to be told when they are "excessively wide".   It is sad but true, that a very small number of AM stations pride themselves on being wide.  They know they are excessively wide and they are proud of it. They see this as a technique to try and keep QRM away from them. I totally disagree with this thought process.   When I tune these stations in, its not unusual to hear them out at least 15 KHz each side of center frequency and sometimes they go out even further.  The buckshot just goes on and on and they could care less how many QSO's they are messing up.  When I check bandwidth, the first thing I do if they are very strong, is to dial in my external RF attenuator to get their signal down to an "S9".  I usually take AM bandwidth measurements with a 6 KHz IF filter.   

One thing for sure Dave, this is something that we just have to live with because these guys will never see it any other way.  In fact, they love to argue their point to death.

Joe, W3GMS
  
        

Some of us have already adopted this solution :P

Stop talking about Carl!   ;D

Don't forget that when you are coming in 40 dB over 9, splatter products 40 dB down (well within FCC specs) will still be audible at S-9.

There are two independent factors regarding the bandwidth of a signal: the frequency range of the audio fed to the modulator, and spurious distortion products.

The apparent bandwidth of a signal is the sum of the actual bandwidth plus the selectivity of the receiver. Many times, hams seem unaware of this; the ultimate example is the occasional SSB operator heard pissing 'n moaning about the "wide" AM carrier producing a squeal all the way across 3 kc/s of the band. A 6 kc/s wide AM signal will have an apparent bandwidth of about 9 kc/s of total dial space on a typical SSB receiver.

Don's very important point is missed by many of the complainers who understand a lot more about griping than they do about measurement.  Those expensive highly selective filters aren't put into spectrum analyzers and frequency selective voltmeters just to take up space but are critical in attempts to measure the true bandwidth of a signal.  Typical filters used in ham receivers and transceivers are relatively broad and have a less than perfect shape factor.