Dear Neil Radio8Z:
Because of your casual observation that your new signal is less scratchy than the old one, we could in all fairness conclude that THEREFORE the new transmitter is in fact more efficient. All things being equal. That is, assuming you are using the same antenna in the same location.
And thank you for understanding my screwy honesty.
I knew this transmitter was more efficient than my other one because I bench tested it before using it on the air so it was not a surprise to me that the perceived improvement in the signal was observed since it was predicted by the bench test. The efficiency of the transmitter was not determined by the listening test.
Neil
"The efficiency of the transmitter was not determined by the listening test."
Measuring efficiency of single digit differences at the power levels involved is incredibly difficult to obtain by the ear alone, especially between levels of 100mW vs 113mW. A .013 difference wont even make the hump on the spec-an change enough to see it either.
But if a range increase is noted, and the power levels are within a reasonable range of each other...such as the .013 difference of 100mW and 113mW, the "quieting" at the outer rim of known acceptable listening range can also be taken as viable evidence of increased performance when the antenna normally used is not altered an any way with the exception of minor peak tuning from one transmitter to the other.
If Carl's test were conducted at an unreasonable power setting..say 120mW or above, then the external antenna test and comparison of unit 1 vs unit 2 would be invalid simply because unit 1 cannot go beyond its designed power level versus unit 2 with a power adjustment pot. But I think Carl's test at 113mW is very reasonable and within range of the established limits to conduct a real world test.
At the power levels involved, every fragment of efficiency and power transfer ratio (matching/coupling) counts. Those with the proper test gear obviously have an advantage to perform accurate measurements..depending upon if the test equipment itself is calibrated correctly for the frequencies involved.
But for the majority, even though I have lab grade test gear good enough to satisfy the field agent of my measurement and compliance logs, I speak for those majority who do not have the luxury of depending on an expensive piece of test gear to tell them if their setup is achieving that single digit efficiency advantage. The next best method for those folks is to take their new transmitter and connect it up to their existing antennas used to get the signal to the audience...be it a window pane antenna, or an outdoor 3 meter rod over a set of ground radials, or even a piece of wire strung up off a tree or flag stick off the side of a wall.
The whole point..in my opinion..is to achieve that single digit performance increase through the antenna system normally used for regular operations. The bench testing gives a good starting reference point, but it does not provide nor prove that the increased efficiency noted on the bench will also show itself through the regular antenna system. The only way to find that out is to run it through said antenna system and measure it with the test gear, or use the alternative method....taking a radio and noting if the outer rim range has improved or not.
There are going to be operators out there who are going to test their new transmitter into their external antenna systems and use basic methods of determining improvement or not simply because that is all they have to work with. In such cases, those who have done the testing in this manner should not be discouraged because it is not accurate. Who cares about the single digit accuracy! I want the improved efficiency to count where it really counts...out there in the real world where the audience is.
It stands to reason that if one notes an improvement, even if that improvement comes from better quieting of the carrier at the outer range limit..well in my book that makes the testing effort worthwhile and credible for others to base their testing.
I do not discourage accuracy or encourage setting aside the use of actual lab testing procedures. However every engineer knows that no matter what the test gear is telling you on the bench, it does NOT tell you what the results will be running through the external antenna system until specific tests are conducted with said external antenna system.
When I design and prototype, I rely on the test gear and bench to be the starting point, the reference maker. After that, its time for the real world testing and application. I never consider the bench/lab tests as the end all be all when the other half of the testing....using the outdoor antenna system....is left out of the loop.
Part 73 transmitter manufacturers have real world outdoor facilities equipped with antenna systems to conduct extended tests of that transmitter on a real antenna in real world settings. They do not just rely on dumping power into a dummy load and call it a day. Only after performing field tests do they slap their label onto the unit and crate it up to ship out. It is part of the overall quality control and quite frankly..if they did not do such extensive real world testing and provide a print out of those tests upon delivery of the transmitter, I would look elsewhere to a manufacturer that does conduct real world testing.
This is why I began to suggest comparing one unit vs another here with the new 5K unit. It is claiming things that result from the most basic setup...a piece of wire hung off the ceiling tiles in a controlled environment. That does not validate any sort of claim to improved performance or efficiency when no test data using real world situations are provided. It is almost a "either it will or it wont" grey area and that makes folks a bit nervous.
The economics these days makes anyone cautious. Although a good deal and fair price, the cost of the unit cannot be justified or meet with my approval to cut a check when there is missing data regarding real world performance.
Despite Carl's "window pane" antenna setup, the testing he has conducted is credible and can be a reference for those who operate under similar conditions. His results are from real world application. If my station was relying on a window pane for an antenna, I would be far more comfortable cutting that check knowing that another user with similar setups did in fact get better results.
Now all we need is someone who has a 3 meter antenna system to submit their findings. I suggest that those who do submit their findings also include as much description about their antenna system as possible so as to give others an idea about what to expect from their own setup, perhaps even get new ideas to improve their antenna system..such as Carl noting about certain things with his antenna..and he got a suggestion to tune out the excess capacitance, and now that same window pane antenna is performing better on both the 3K unit and the 5K unit!
What better proof does anyone need that real world performance testing is worthwhile?!
RFB
Neil,
Your transmitter, as you certainly know, is a curiosity to the brethren. But it is personal intellectual property of yours.
But we do naturally wonder if the design will later be made available for observation.
Does it have output power control? That will be the new "cincher" for future adopters.
Actually, the 20 pF coupling capacitor in my comparison test between AMT3000 & 5000 was only used on the 5000, based on the reasoning that:
Previously the 3000 had loaded into the antenna and could be peak tuned, whereas the 5000 could only be loaded into the antenna with the 20 pF cap and then peak tuned.
What I will try is a test in which both transmitters load via the 20 pF capacitor and the AMT5000 is set at 100 mW. Of course the power output of the AMT3000 is not adjustable, and is 100 mW according the Specifications, page 35, AMT3000 Manual.
Getting ready to re-test the AMT5000 with input to final adjusted closer to the target setting (100 mW).
It is windy today, and this must be affecting the capacitances of surrounding materials, because the reading is floating around at occasional times, of course with the audio silent, whereas the current normally finds itself and stays constant.
The AMT5000 Manual features a wonderful Equivalence Table on page39 showing a vast set of current/voltage settings that result in 100mW, and this got things going very quickly, combined with the fact things were already set at 113 mW. At first try I got 95 mW, next 110 mW, then settled on 102 mW, which I thought reasonable given the unstable atmosphere with regard to capacitance, which of course causes a changing milliWatt condition.
Shortly I will do the drive and observe the range with this setup.
Carl,
Continue field testing and reporting. Your results will be important. As discussed earlier, you need not be "laboratory perfect" with this, just describe your test conditions and observations.
Carl commented: "Your transmitter, as you certainly know, is a curiosity to the brethren. But it is personal intellectual property of yours. But we do naturally wonder if the design will later be made available for observation. Does it have output power control? That will be the new "cincher" for future adopters."
Yes, it does have a power control. I do intend to share the schematic and test data on this when I can do a decent write up on the project. I can't promise when but it could be soon I hope. I hate to tease like this but this is on my to do list.
Neil
Thank you Neil Radio8Z for your generosity to the cause of low power radio.
I am just headed out the door to do this next test with AMT5000.
And I am also looking forward to the modification to Big Talker to convert it into 1-Watt long wave.
There is never a rush on these things, it is always good to have things coming in the future.
Did the latest AMT5000 distance test in the present series, and while putting the data on paper I found a probable math error in the first AMT5K test which may actually have reached farther than reported.
I'll make excuses later for my poor math, but tonight I will gather the data and corrections in what I expect will be an accurate report on the 2 AMT5K distance tests.
In studying the procedure I'm using to measure the transmitters I have spotted several ways of improving the strategy for more accurate results.
I have a Stanley Measuring Distance Tool with handle and wheels, so I can be very precise in feet.
It also occurred to me that the distances in my estimates were calculated along the sidewalk, but my car and radio are listening from out in the street, which is X-Number of feet farther away.
Yet, I WILL post today's test and the corrections from the last AMT5000 test right here on this thread.
Then I will start from scratch with new improved testing methods and a whole new blog thread.
The first tests have been a test of the testing procedure.
This evening I poured over my test notes and find them confusing. That's a bad thing in science. Tests must be replicable and the documentation must be clear.
I'm taking charge here, and here's the plan...
Today's re-testing of the AMT5000 will amount to this simple anecdote:
"With the input power turned down from 113 mW to 105 mW it didn't seem to go as far. Instead of reaching the edge of the property of the 8th house west of here, it only made it to the middle of that 8th lot." -end anecdote.
The newly announced plan is that I will check and double check every step of the way as never before. The next comparison test between the AMT3000 and the AMT5000 will be flawless.
And I know why I have been slightly scrambled.... it's the solder smoke from building the kit! I went temporarily insane for a little while.
Sounds like you are making good progress Carl.
RFB
Carl,
You are finding that the best way to learn science and engineering testing is to do it. In general, if you describe the test methods and conditions then the results are useful to others even if they choose not to replicate them.
One of my favorite questions to pose to my students when I taught laboratory methods in engineering colleges (three of them) was to write on their reports "Which voltmeter did you use?" The labs had up to 24 meters on the shelf so they couldn't answer the question. Then I would tell them that if there was a question regarding their data they had no way to confirm that the instrumentation was calibrated properly. The lesson was to always record the serial number of any instruments used.
No, you don't have to do this but it is one of the principles of good engineering practice and I only mention it to illustrate that these things are not as simple as they seem until one does it.
Anyway, I am enjoying with interest your reports.
I have a hiker's GPS which gives the distance between two points which would be handy for these range measurements. Your distance wheel is good but it is probably hard to traverse a straight line. I have this funny picture in my mind of you driving across neighbors' lawns with your door open and you holding the measuring wheel to the ground.
Neil
I may not have driven up on neighbor's lawns but the auto portion of the trip did involve trash cans in the street, parked cars, rushing cars coming and going and me trying to count the houses on my left which corresponded with the wheel measurements taken earlier.
The Stanley measuring wheel was rolled along the side walk and at one time a man walking a giant dog was headed toward me so I veered a few feet down a paved driveway to give him room, therefore there's a "driveway bubble" in my straight line.
And the straight line is a curve on the east/west direction, because these streets are all curvy. I think my south/north line down my driveway is straight.
Topography, which I think refers to the hills and valleys, is a missing factor in the measurements. I hold the secret belief this area was settled right on top of a field of indian burial mounds, it's uphill in all directions from here, and the undulating upping and downing goes on for a few miles in all directions.
One of my many postmen said the reason we have so many different mail men is because no one wants this route because of all the ups and downs.
I'll be sure to add the serial number of my meter!
You could also just measure one of the sidewalk sections, which they will all be of the same size in the straight lane and merely count them then multiply. Add the distance from the TX point to the sidewalk to the total and there ya go.
Not exactly precise but I doubt that without precision techniques and precision equipment to do the testing, the results are not going to be precise down to the nano-fraction inch.
IMO..you already demonstrated the 5K does deliver where it counts with the improved efficiency into your real world antenna system. I think that any further attempts at precision here will only result in discovering the 20 ways to Doom when all it takes is 1.
RFB